Blood Pressure Category Calculator: What Three Sets of Criteria Call the Same Reading

Where one adult blood pressure reading falls on the 2025 AHA/ACC criteria, the 2024 ESC criteria and the 2023 ESH criteria at the same time — shown side by side rather than resolved, because a systolic of 130 to 139 is stage 1 hypertension in the United States and is not hypertension at all in Europe, and a reader told only one of those answers has been told the least useful half of the truth. Classified on the worse of the two readings, as all three documents instruct. One reading taken once is not a diagnosis of anything; a reading at or above 180 systolic or 120 diastolic is the one figure here worth acting on today.

Last updated 5 Oct 2026

This sorts a measurement into a published category. It is not a diagnosis, and one reading taken once is rarely enough to act on. If a result here worries you, take it to a doctor rather than to the internet.

One reading, three classifications, and the disagreement between them

systolic, diastolic and age -> where the reading falls on all three sets of criteria at once

Systolic — the TOP number (mmHg)The higher of the two numbers, measured as the heart contracts. Enter the reading you actually got, not an average you have adjusted in your head. The page refuses outside 50 to 300 mmHg, and refuses any pair where the top number is less than 5 mmHg above the bottom one, because that is a measurement artefact rather than a blood pressure. Readings in children and adolescents are not interpreted here at all: hypertension under 18 is defined against percentiles for age, sex and height, and the paediatric blood pressure percentile page in the medical set does that properly.

Diastolic — the BOTTOM number (mmHg)The lower of the two numbers, measured between beats. This one decides the category more often than readers expect — both the American and the European documents classify on whichever of the two numbers is worse, and a calculator that looks only at the top number will tell a 28-year-old with 126/94 that everything is fine. The rows below say which of your two readings set the category.

Age in years — adults onlyUsed for two things only: to refuse the calculation under 18, and to add the notes that are specific to younger and to older adults. It does not change which category a reading falls in — none of the three sets of adult criteria is age-adjusted, which is itself a thing worth knowing, since the same 150/85 means something different at 30 and at 80 and no classification captures that.

5of 9Example

a 52-year-old adult with a single reading of 134/86 mmHg

The worse-of-the-two rule, three times over, and how nine joint positions fall out of it

For each set of criteria: category = max(category from systolic, category from diastolic) · AHA/ACC 2017 and 2025: systolic steps at 120, 130, 140; diastolic steps at 80, 90 · ESC 2024: systolic steps at 120, 140; diastolic steps at 70, 90 · ESH 2023: systolic steps at 120, 130, 140, 160, 180; diastolic steps at 80, 85, 90, 100, 110 · this page’s own care-today flag: systolic ≥ 180 or diastolic ≥ 120 · pulse pressure = S − D · mean arterial pressure ≈ (S + 2D) / 3

S, D: the systolic and diastolic readings in mmHg, as measured, not averaged or adjusted
max(): the worse-of-the-two rule, and the worse-of-the-two rule, which is not this page’s invention: the 2017 AHA/ACC guideline instructs that a person whose systolic and diastolic fall in two categories is assigned to the higher one, and the 2023 ESH criteria state that the category is defined by the highest level of blood pressure, whether systolic or diastolic
1 to 9: why there are nine of them: the three sets of criteria have different numbers of bands (four, three and six) and draw their lines in different places, so there are more distinct JOINT positions than any one of them has bands. Sweeping every whole-mmHg pair from 60/30 to 300/200 produces exactly nine distinct combinations of the three verdicts, and they turn out to be totally ordered — each of the three verdicts is non-decreasing as you move up the list. That is why this page can show a single number: the nine steps carry all three classifications without inventing a correspondence between them
one reading: this page takes ONE reading and does not average: this page takes one reading. It does not average, because averaging two readings you typed is arithmetic you can do yourself, and because what the criteria ask for is an average of at least two readings on at least two separate OCCASIONS, which a single page visit cannot represent
180, 120: the published wording is “higher than 180 and/or higher than 120”. This page tests at or above 180 and at or above 120, one mmHg more cautious in each limb, deliberately

Worked example

a 52-year-old adult with a single reading of 134/86 mmHg
Classify each number separately, three times, before combining anything. Systolic 134: on AHA/ACC it is in 130 to 139, which is their stage 1 step (3 of 4). On the 2024 ESC criteria it is in 120 to 139, their elevated step (1 of 2). On the 2023 ESH criteria it is in 130 to 139, their high-normal grade (2 of 5). Diastolic 86: on AHA/ACC it is in 80 to 89, also their stage 1 step (3 of 4). On ESC it is in 70 to 89, their elevated step (1 of 2). On ESH it is in 85 to 89, their high-normal grade (2 of 5).
Take the worse of the two limbs for each set of criteria, which is what each of them instructs. AHA/ACC: max(3, 3) = 3, STAGE 1 HYPERTENSION. ESC 2024: max(1, 1) = 1, ELEVATED BLOOD PRESSURE — which the 2024 ESC document is explicit is not hypertension. ESH 2023: max(2, 2) = 2, HIGH-NORMAL. One reading, three answers, and only one of them contains the word hypertension.
The joint position is step 5 of 9. That is what the headline number above means: the fifth of the nine distinct combinations of the three verdicts that exist across the whole plausible range of readings. It is not a score, it is not out of ten, and a higher number is not a worse person — it is an index into a list of nine label-combinations, printed as a number only because the engine behind this page carries one number and the chip beside it carries the words.
Now the thing readers most want and this page refuses to give. Which of the three is right? None of them is a measurement and all three are defensible. The 2017 AHA/ACC guideline lowered the American threshold to 130/80 on the grounds that risk rises continuously from well below 140/90 and that labelling earlier gets people to lifestyle change and to risk assessment sooner. The 2024 ESC criteria kept hypertension at 140/90 on the grounds that the label should mark where drug treatment is warranted for most people, and created ELEVATED BLOOD PRESSURE (120 to 139 or 70 to 89) to cover the range where treatment benefit exists but average risk does not justify treating everyone. Those are two different jobs for one word. Nothing in your arteries changes at 130, or at 140.
How far this reading is from the numbers that matter. Systolic 134 is 6 mmHg below 140; diastolic 86 is 4 mmHg below 90, so the governing limb is 4 mmHg below the threshold all three sets of criteria share. It is 46 mmHg and 34 mmHg respectively below this page's care-today flag, so the nearer limb is 34 mmHg clear of it. Pulse pressure is 134 − 86 = 48 mmHg. Mean arterial pressure by the usual approximation is (134 + 172) / 3 = 102 mmHg.
What one mmHg does, and why that should make you distrust the whole exercise. At 139/89 this page reads step 5 and only the American document uses the word hypertension; at 140/89 it reads step 6 and all three do. One millimetre of mercury on the systolic limb moves a reader across the word. Going the other way, 125/84 is step 4 and 125/85 is step 5, because the 2023 ESH high-normal grade begins at a diastolic of exactly 85. From this example's own 134/86, by contrast, the diastolic has to fall to 84 AND the systolic to 129 together before any of the three verdicts changes at all — because at a systolic of 134 the ESH limb is already high-normal on its own, so lowering the diastolic alone changes nothing. That is the worse-of-the-two rule doing exactly what it is for, and it is the reason a page classifying on the top number alone gets different answers from this one. Change the diastolic from 86 to 90 instead and all three say hypertension, which is step 6. The whole of that range — from "normal" to "hypertension" on three different documents — is about 11 mmHg wide, and the published error on a single measurement is comparable to it: a regular cuff on an arm needing an extra-large one overstated systolic pressure by 19.5 mmHg on average in a randomised trial. The honest reading of this page is therefore that the step number is far less certain than it looks, which is exactly why every one of these documents asks for repeated readings on separate occasions.
What this page will not tell you. Whether to take a tablet, whether your pressure is controlled, what your cardiovascular risk is, or anything at all about a reading taken in pregnancy or in anyone under 18. It reports where one number pair lands on three published classifications, says where they disagree, and flags the one reading that is worth acting on today.

The nine joint positions, with an example reading for each

Nine rows, not four, three or six, because the three documents put their lines in different places and the joint position carries more information than any one of them. Sweeping every whole-mmHg pair this page accepts — systolic 50 to 300, diastolic 20 to 200, with the systolic at least 5 mmHg above the diastolic, which is 33,341 pairs — yields exactly these nine combinations and no others, and they are totally ordered: each of the three verdicts is non-decreasing down the column. That is a property of these particular criteria rather than a design choice, and it is checked by this page’s test suite against an independently written implementation of all three classifications. The bold cells are the disagreements. Steps 4 and 5 are the ones that matter: an adult told “you have stage 1 hypertension” in Delhi or Dallas and “your blood pressure is high-normal” in Düsseldorf has had the same reading classified by two documents that read the same evidence.
Step	Example reading	AHA/ACC 2017 and 2025	ESC 2024	ESH 2023
1	112/66	Normal	Non-elevated	Optimal
2	112/74	Normal	Elevated BP	Optimal
3	124/76	Elevated	Elevated BP	Normal
4	128/82	Stage 1 hypertension	Elevated BP	Normal
5	134/86	Stage 1 hypertension	Elevated BP	High-normal
6	146/88	Stage 2 hypertension	Hypertension	Grade 1 hypertension
7	166/94	Stage 2 hypertension	Hypertension	Grade 2 hypertension
8	172/112	Stage 2 hypertension	Hypertension	Grade 3 hypertension
9	186/118	Stage 2, severe range	Hypertension, immediate-treatment range	Grade 3 hypertension

Where each set of criteria draws its lines, in mmHg

This is a comparison table assembled for this page, not a reproduction of any one guideline’s own table — no guideline publishes a column for the other two. Each figure is cited to its source in the references. Note the two rows in bold, which are the whole story: AHA/ACC moved the hypertension threshold to 130/80 in 2017 and kept it there in 2025, while both European documents left it at 140/90; and the 2024 ESC criteria lowered the bottom of their first concerning band to a diastolic of 70, which neither of the others did. Note also the last row: all three instruct that the category is set by whichever of the two readings is worse, so classifying on the systolic alone is wrong on every one of them.
Threshold	AHA/ACC 2017 and 2025	ESC 2024	ESH 2023
Top of the lowest band, systolic	under 120	under 120	under 120
Top of the lowest band, diastolic	under 80	under 70	under 80
Where hypertension begins, systolic	130	140	140
Where hypertension begins, diastolic	80	90	90
Number of named bands	4	3	6 plus two isolated patterns
Highest named band begins at	140/90 (stage 2)	140/90 (hypertension)	180/110 (grade 3)
Severe / immediate-treatment marker	above 180 and/or above 120	at or above 180/110	grade 3 is at or above 180/110
Readings needed before a diagnosis	average of ≥2 readings on ≥2 occasions	out-of-office confirmation preferred	at least 2 office visits within 4 weeks, unless grade 3 or organ damage
Which limb decides the category	the higher of the two	the higher of the two	the higher of the two

What a proper measurement requires, and what each shortcut costs

The reason this table is on a calculator page: the arithmetic above is exact and the input is not. Every one of the nine steps is separated from its neighbour by between 5 and 20 mmHg, and the measurement errors in this table are the same size. A page that printed a category from one reading without saying this would be dressing up a guess.
What the method assumes	Why	What is known about the cost of skipping it
A cuff that fits the arm	The cuff bladder has to encircle most of the upper arm for the pressure to be transmitted to the artery	The largest measured error of any item here. In a randomised crossover trial of 195 adults, using a regular-size cuff on people who needed a LARGE cuff overstated systolic pressure by 4.8 mmHg; on people who needed an EXTRA-LARGE cuff it overstated it by 19.5 mmHg; and on people who needed a SMALL cuff it understated it by 3.6 mmHg. Nineteen and a half mmHg is wider than the entire gap between “normal” and “hypertension” on any of the three documents.
At least five minutes sitting quietly first, back supported, both feet flat on the floor, legs uncrossed	Blood pressure falls as you settle and rises with isometric effort such as crossed legs or an unsupported back	Both the published measurement advice and all three sets of criteria build this in; a reading taken on arrival is not the reading the thresholds were derived against
The arm resting at chest height on a table, cuff on bare skin	Below heart level the hydrostatic column adds pressure; a sleeve under the cuff adds artefact	Arm position alone can move a reading by several mmHg in either direction
No food, drink, caffeine or tobacco for 30 minutes, and an empty bladder	All of these raise blood pressure acutely and for longer than people expect	A single cigarette or coffee can be worth more than the width of a band on this page
No talking during the measurement	Speaking raises pressure immediately	Advice given in all published measurement protocols
At least two readings, a minute or two apart, averaged	Consecutive readings in the same person differ by several mmHg from beat-to-beat and respiratory variation alone	Both American and European criteria specify an average rather than a single reading. This page takes one reading, which is the page’s own limitation and is stated here rather than buried
Readings on more than one occasion, preferably outside a clinic	A single office reading has poor specificity for sustained hypertension; white-coat and masked patterns both exist and both mislead in opposite directions	AHA/ACC requires an average of at least two readings on at least two occasions; ESH requires at least two office visits within four weeks unless the reading is grade 3 or there is organ damage; the 2024 ESC criteria prefer out-of-office confirmation outright

Where each figure comes from, and what was rejected

The recurring theme: the numbers are facts and are used freely; the tables and the prose belong to the bodies that wrote them and are not reproduced. One departure from a source is recorded above rather than buried — the care-today flag is set one mmHg more cautious than the published wording, deliberately. And one thing could not be sourced at any price: a classification on which all three documents agree. It does not exist, which is the whole reason this page shows three columns and settles nothing.
Figure	Source used	What was rejected, or could not be established
Normal, elevated, stage 1 and stage 2, and the 130/80 threshold	The 2017 ACC/AHA guideline (Whelton and colleagues, Hypertension 2018;71:e13–e115), unchanged in the 2025 AHA/ACC guideline (Jones and colleagues, 2025). Cited as figures and by the guideline’s own instruction on mixed categories, which is quoted in one sentence. The United States Government route for the same numbers is the CDC’s high blood pressure page, which prints normal as under 120 and under 80, elevated as 120–129 with a diastolic under 80, and high blood pressure as 130 or higher or 80 or higher — and which attributes them to the ACC and the AHA.	Reproducing either guideline’s classification table as a table. The guidelines are the work of private non-profit bodies and this site carries advertising; the numeric thresholds are facts and are used as such, but no table, criteria list or passage of guideline text is reproduced. The CDC page does NOT carry the stage 1 / stage 2 split or any crisis threshold, so those are cited to the guideline itself as figures.
The 2024 ESC categories: non-elevated under 120 and under 70, elevated 120–139 or 70–89, hypertension 140 or above or 90 or above	The 2024 ESC Guidelines for the management of elevated blood pressure and hypertension (McEvoy and colleagues, Eur Heart J 2024;45(38):3912–4018). Cited by figure. The diastolic floor of 70 for the elevated band is the change most readers have not heard of and is the reason step 2 exists on this page.	Any suggestion that the ESC replaced the ESH criteria. They are different bodies and the 2023 ESH criteria remain published; this page shows both rather than assuming the newer document supersedes the older one, because that is a judgement about professional standing and not a fact.
The 2023 ESH grades: optimal, normal, high-normal, grades 1 to 3, and the isolated systolic and isolated diastolic patterns	The 2023 ESH Guidelines for the management of arterial hypertension (Mancia, Kreutz and colleagues, J Hypertens 2023;41:1874–2071; DOI 10.1097/HJH.0000000000003480). Cited by figure, including its own statement that the category is defined by the highest level of blood pressure, whether systolic or diastolic.	Reproducing its classification table. Only the numeric thresholds and the worse-of-the-two rule are used.
The worse-of-the-two-readings rule	Stated in both families. The 2017 ACC/AHA guideline instructs that individuals whose systolic and diastolic fall in two categories are assigned to the higher category; the 2023 ESH criteria state that the category is defined by the highest level of blood pressure, whether systolic or diastolic. Implemented as max(systolic step, diastolic step) separately within each set of criteria.	Classifying on the systolic alone, which is what a large number of online blood pressure calculators do and which systematically under-reads the isolated-diastolic pattern common in younger adults. Also rejected: taking the max ACROSS criteria sets, which would be meaningless, since their steps are not the same quantity.
The care-today flag at 180 or 120	The published severe-hypertension and hypertensive-emergency threshold, which the AHA states as higher than 180 and/or higher than 120 mmHg, and which the 2025 AHA/ACC guideline states as blood pressure above 180/120 mmHg. The symptom list that turns it into an emergency — chest pain, breathlessness, back pain, numbness or weakness, change in vision, difficulty speaking — is the published one. The repeat-after-a-minute advice for a reading without symptoms is the published advice.	The strict inequality. The brief for this page gave the threshold as “at or above 180 and/or at or above 120”; the published wording is “HIGHER THAN 180 and/or HIGHER THAN 120”, so on the guideline’s own words a reading of exactly 180/120 is not above it. This page nonetheless tests at or above 180 and at or above 120, one mmHg more cautious in each limb, because the cost of being one mmHg early here is a wasted phone call and the cost of being one mmHg late is not. The departure is stated on the page rather than hidden. Also noted: the AHA’s own public guidance now says to wait ONE minute before repeating, not the five minutes that older versions of that advice gave.
Cuff size errors of 3.6, 4.8 and 19.5 mmHg	Ishigami J, Charleston J, Miller ER and colleagues, the Cuff(SZ) randomised crossover trial, JAMA Intern Med 2023 (published online 7 August 2023), 195 participants. Cited by figure.	Nothing. This is the single most useful number on the page for a reader using a home monitor, and it is a measured randomised-trial result rather than an expert opinion.
Measurement technique	The CDC’s page on measuring blood pressure, a United States Government work: at least five minutes seated with the back supported, feet flat and legs uncrossed, nothing to eat or drink for 30 minutes beforehand, bladder emptied, arm resting on a table at chest height, cuff against bare skin and snug but not tight, no talking, and at least two readings one or two minutes apart.	The UK national institute’s guidance on hypertension diagnosis and the 135/85 home-monitoring threshold it uses, which would have been directly relevant. Its open content licence is United Kingdom-only and forbids displaying the licensed information next to advertising, and this site carries advertising, so none of it is used anywhere on this site. That is a licensing exclusion and not a judgement about the guidance.
Hypertension prevalence in India of 35.5%	The ICMR-INDIAB cross-sectional survey of 113,043 adults (33,537 urban and 79,506 rural) across 31 states and union territories, surveyed 2008 to 2020 and published in The Lancet Diabetes & Endocrinology in 2023. Also from the same survey: diabetes 11.4%, prediabetes 15.3%, generalised obesity 28.6%, abdominal obesity 39.5%, dyslipidaemia 81.2%.	Treating the figure as a reason to interpret a reading differently. It is context for how common these readings are, nothing more.
Paediatric readings	Not interpreted here at all. Hypertension under 18 is defined against percentiles of age, sex and height, with absolute limbs layered on top.	Any attempt to apply adult bands to a child, which would be wrong in both directions depending on the child’s height centile. The paediatric blood pressure percentile page in the medical set does this properly and this page refuses under 18.
Pregnancy	Not interpreted here either. The band notes flag it.	Applying these bands in pregnancy. The thresholds that matter in pregnancy are different and lower, severe is defined at 160/110, and a reading of 140/90 is already actionable. A page that applied adult non-pregnant bands to a pregnant reader would under-read her, which is the dangerous direction.
A single-number risk estimate	Not offered. Nothing on this page estimates cardiovascular risk.	Every published cardiovascular risk equation this page’s author looked at: they need lipids, smoking status, diabetes status and often more, they are population- and region-specific, and most are published by bodies whose material this site cannot reproduce. Printing a risk percentage from two blood pressure numbers would be the worst kind of plausible-but-wrong answer.

Why one reading gets three different names, and why this page refuses to pick one

The most confusing thing about blood pressure numbers is not the numbers. It is that two groups of cardiologists, reading the same trials, drew the line in different places and did not change their minds. In 2017 the American College of Cardiology and the American Heart Association moved the definition of hypertension in adults down from 140/90 to 130/80. The 2025 AHA/ACC guideline kept it there. Neither European document followed. The 2024 ESC guidelines on elevated blood pressure and hypertension keep hypertension at 140/90 and the 2023 ESH guidelines do the same. So a 134/86 reading is stage 1 hypertension on American criteria, elevated blood pressure and explicitly not hypertension on the 2024 ESC criteria, and high-normal on the 2023 ESH criteria. All three of those are correct. A calculator that gives you one of them and stops has given you the least useful part of the answer, which is why this page gives all three and marks the places where they disagree.

What the disagreement is actually about, since it is not about evidence. Blood pressure and cardiovascular risk are related by a smooth curve with no step in it. Risk is higher at 135 than at 125 and higher at 125 than at 115, continuously, all the way down into the range nobody calls high. There is no number at which anything changes. So a threshold is not a discovery; it is a decision about what a word is for. The American position is that the word should mark where risk has risen enough to be worth naming and acting on, which gets people assessed and counselled earlier, and the 2017 guideline is explicit that most people newly labelled by the change were not expected to need medication. The European position is that the word should mark where drug treatment is warranted for most people, so that the label carries a treatment implication — and the 2024 ESC criteria then invented a middle category, elevated blood pressure (120 to 139 systolic or 70 to 89 diastolic), precisely to cover the range where treatment benefit exists in trials but average risk does not justify treating everyone in it. Both are coherent. Neither is a measurement. Nothing in anybody’s arteries happens at 130, or at 140.

The second disagreement, which almost nobody has heard of, is at the bottom. The 2024 ESC criteria set the top of their lowest band at a diastolic of 70, not 80. So a reading of 115/76 is normal on American criteria, optimal on the 2023 ESH criteria, and elevated blood pressure on the 2024 ESC criteria. That is step 2 of the nine on this page, and it exists only because of that one change. It is in the table above because a reader who looks up “elevated blood pressure” after an ESC-based consultation and finds an American definition will not understand what happened to them.

Which of your two numbers decides the category, and why getting this wrong is the commonest error in the arithmetic. All three documents say the same thing: classify on whichever of the two readings is worse. The 2017 ACC/AHA guideline instructs that a person whose systolic and diastolic fall in two categories is assigned to the higher one. The 2023 ESH criteria state that the category is defined by the highest level of blood pressure, whether systolic or diastolic. This page implements that as the maximum of the systolic step and the diastolic step, computed separately inside each set of criteria, and it prints which of your two readings governed. A great many online calculators classify on the systolic alone. The cost of that is not random: it falls on younger adults, in whom a raised diastolic with a normal systolic — the isolated-diastolic pattern — is the usual presentation. A 28-year-old at 126/94 is hypertensive on all three sets of criteria and a systolic-only calculator will tell her she is fine.

What the headline number is, and what it is not. It is an index into a list of nine, not a score. The three documents have four, three and six named bands and they do not line up, so the number of distinct JOINT positions is larger than any one of them. Sweeping every whole-mmHg pair this page accepts — systolic 50 to 300, diastolic 20 to 200, with the systolic at least 5 mmHg above the diastolic, which is 33,341 pairs — produces exactly nine combinations of the three verdicts, and those nine turn out to be totally ordered, so one number can carry all three classifications without inventing any correspondence between them. Step 7 is not “worse than” step 5 by two units of anything. It is the seventh entry in a list. The words in the chip beside it are the output; the number is how the page got there.

One reading is not a diagnosis, and this page takes one reading. That is worth stating as plainly as possible, because it is the realistic failure mode: someone takes a home reading after an argument, gets 148/96, finds a web page, and believes a thing about themselves. Every one of the three documents asks for more than this page has. AHA/ACC asks for an average of at least two readings taken on at least two separate occasions and recommends out-of-office measurement to confirm. The 2023 ESH criteria ask for at least two office visits within four weeks, unless the reading is in the highest grade or there is already organ damage or established cardiovascular disease, in which case waiting is the greater risk. The 2024 ESC criteria go further and prefer out-of-office confirmation outright, because a single office reading has poor specificity for sustained hypertension. There are two named ways that goes wrong in opposite directions: white-coat hypertension, where clinic readings are high and home readings are not, and masked hypertension, where the reverse is true and the person who looks fine in a clinic is not. A single reading cannot tell those apart from the real thing, and neither can this page.

And the measurement itself is much worse than the arithmetic. The nine steps on this page are separated by 5 to 20 mmHg. The published errors from ordinary measurement shortcuts are the same size or larger. The one that costs most, and is checked least, is cuff size: in a randomised crossover trial of 195 adults, using a regular cuff on an arm that needed an extra-large one overstated systolic pressure by 19.5 mmHg on average — wider than the whole distance from “normal” to “hypertension” on any of the three documents. A regular cuff on an arm needing a large one overstated by 4.8 mmHg; on an arm needing a small one it understated by 3.6 mmHg. Add to that the five minutes of sitting still that the thresholds assume, the 30 minutes without caffeine or tobacco, the emptied bladder, the supported back, the uncrossed legs, the arm at chest height and the not talking, and it becomes clear that a category computed from one casual reading is a category computed from a number that could easily be in a different band. Measure it properly, measure it more than once, and take the series to a doctor.

What this page deliberately does not do. It does not interpret a reading in anyone under 18 and refuses to try: blood pressure in children and adolescents is defined against percentiles of age, sex and height, and the paediatric blood pressure percentile page in the medical set does that properly. It does not interpret a reading in pregnancy, where the numbers that matter are different and lower. It does not estimate cardiovascular risk, because every published risk equation needs lipids, smoking and diabetes status at a minimum, and printing a risk percentage from two numbers would be exactly the kind of confident wrong answer this page exists to avoid. It does not tell you whether to take medication, or whether your treatment is working. It does not average readings for you. And it carries no advice on alcohol or tobacco beyond pointing at the two pages in this set that count them — the alcohol units page and the pack-years page — because this page’s job is to classify a number honestly, not to tell you how to live.

Frequently asked questions

My reading is 135/85. Do I have high blood pressure or not?

Both, depending on whose criteria you read, and that is a real answer rather than a dodge. On the 2025 AHA/ACC criteria, 135 systolic is in the 130 to 139 band, so this is stage 1 hypertension. On the 2024 ESC criteria it is elevated blood pressure, a category the ESC document is explicit is not hypertension, because the ESC keeps hypertension at 140/90. On the 2023 ESH criteria, 135/85 is high-normal, the grade immediately below grade 1 hypertension. Three documents, three answers, one reading, and no experiment can settle it, because the disagreement is about what the word hypertension should be used to mark, not about what 135/85 does to a person. What all three agree on is that risk at 135/85 is higher than at 115/70, that the first things to change are not tablets, and that whether medication is worth taking depends on your overall cardiovascular risk and not on the reading by itself. Also: one reading establishes nothing. Take several, properly, over a couple of weeks, and take the numbers to a doctor.

Which number matters more, the top one or the bottom one?

For deciding the category, whichever of the two is worse — that is not this page’s opinion, it is what the guidelines instruct. The 2017 ACC/AHA guideline says a person whose systolic and diastolic fall in two different categories is assigned to the higher category. The 2023 ESH criteria say the category is defined by the highest level of blood pressure, whether systolic or diastolic. So a reading of 126/94 is hypertensive on all three sets of criteria even though the top number looks fine, and this page says so, and prints which of your two readings governed. Beyond the category, the two numbers carry different information. A raised systolic with a normal diastolic — the isolated-systolic pattern — becomes more common with age as the large arteries stiffen and is not a milder condition than having both raised. A raised diastolic with a normal systolic is more common in younger adults and is the pattern most often missed, because calculators and casual attention both go to the first number.

What should I do if my reading is above 180 or above 120?

If you also have chest pain, breathlessness, back pain, weakness or numbness down one side, trouble speaking, a sudden change in vision, confusion or a severe headache, treat it as an emergency: call an ambulance or get to an emergency department now, and do not drive yourself. That combination is a hypertensive emergency and the concern is damage happening at the time to the brain, heart, kidneys or eyes. If you have none of those symptoms, the published advice is to sit quietly, wait a minute or two and measure again, because a single reading this high is very often a measurement problem — wrong cuff size alone was worth nearly 20 mmHg of overestimation in a randomised trial. If the repeat is still at or above 180 systolic or 120 diastolic, contact a doctor today. Severe hypertension with no symptoms is not an emergency department problem, but it is also not a next-month problem. One point of exactness: the published wording is “higher than 180 and/or higher than 120”, so on those words exactly 180/120 does not cross the line. This page flags at 180 or above and 120 or above anyway, one mmHg more cautious, and says so rather than quietly shifting a threshold.

Why does the headline say something like “5 of 9”? Is that a score?

No, and it is worth being clear about it. It is a position in a list of nine, and the list exists because the three sets of criteria do not line up. AHA/ACC has four named bands, the 2024 ESC criteria have three, the 2023 ESH criteria have six plus two isolated patterns, and their boundaries sit at different numbers — AHA/ACC steps at 120, 130 and 140 systolic, the ESC at 120 and 140, the ESH at 120, 130, 140, 160 and 180. Sweeping every whole-mmHg pair this page accepts, 33,341 of them, produces exactly nine distinct combinations of the three verdicts, and those nine happen to be totally ordered, so a single number can index all three classifications at once without inventing a correspondence between scales that do not correspond. The words in the flag beside the number are the actual output. The number is not out of ten, it is not a risk score, and the distance between step 5 and step 7 is not two units of anything.

Can I use this for a child, or during pregnancy?

No, in both cases, and the page refuses under 18 rather than producing a number. In children and adolescents, hypertension is defined against percentiles for age, sex and height, with absolute thresholds layered on top for older children — the same 118/76 can be well within normal for a tall fifteen-year-old and clearly raised for a short seven-year-old, so adult bands are wrong in both directions depending on the child. The paediatric blood pressure percentile page in the medical set does that calculation properly. In pregnancy the numbers that matter are different and lower: 140/90 is already actionable and 160/110 is treated as severe, so applying the bands on this page to a pregnant reader would under-read her, which is the dangerous direction to be wrong in. If you are pregnant and have a reading at or above 140/90, contact your midwife or maternity unit rather than a calculator.

Why doesn’t the page average my readings, or ask for more than one?

Two reasons, and the second is the honest one. First, averaging numbers you have typed is arithmetic you can do without this page. Second, and more importantly, what the criteria actually ask for is not an average of several readings in one sitting but an average across separate OCCASIONS — AHA/ACC asks for at least two readings on at least two occasions, the 2023 ESH criteria for at least two office visits within four weeks, and the 2024 ESC criteria prefer out-of-office confirmation outright. A single page visit cannot represent that, and a page that accepted six readings and printed an average would look as though it had satisfied the requirement when it had not. So this page takes one reading, classifies it honestly, and says in several places that one reading is not a diagnosis. That is a limitation stated rather than designed around.

Is the 2024 ESC guideline not simply the up-to-date one that replaces the 2023 ESH guideline?

They are two different organisations, and deciding that one supersedes the other is a judgement about professional standing rather than a fact, so this page does not make it. The European Society of Cardiology and the European Society of Hypertension are separate bodies; they produced joint guidelines in 2018, then the ESH published its own in 2023 and the ESC published its own in 2024. Both remain published, and they differ in structure: the ESH kept the long-standing seven-row classification with grades 1 to 3, while the ESC collapsed it to three categories and introduced the elevated blood pressure band with its diastolic floor of 70. They agree that hypertension begins at 140/90. Where they differ, this page shows both columns. The same reasoning applies to the American pair: the 2025 AHA/ACC guideline replaced the 2017 one by the same committee process, and it kept the 2017 categories unchanged, so there is nothing to show twice.

My home monitor and my doctor’s reading disagree by 15 mmHg. Which should I believe?

Possibly neither, and the gap is itself the useful information. Three ordinary explanations, and they are not mutually exclusive. Technique: the published measurement requirements — five minutes sitting with the back supported and feet flat, nothing to eat, drink, smoke or caffeinate for 30 minutes, bladder empty, arm at chest height on a table, cuff on bare skin, no talking — are frequently met at home and frequently not met in a busy clinic, and sometimes the reverse. Cuff size: a regular cuff on an arm that needs an extra-large one overstated systolic pressure by 19.5 mmHg in a randomised crossover trial of 195 people, which on its own accounts for a 15 mmHg discrepancy. And the white-coat and masked patterns: some people are genuinely higher in a clinic than out of it and some are genuinely lower, which is the reason all three sets of criteria prefer out-of-office measurement for confirmation. The way through is a validated upper-arm device with the right cuff, a week of readings taken properly at the same times of day, and showing the whole series to a doctor rather than picking the reading you prefer.

How common is a reading in the hypertensive range?

Very. The ICMR-INDIAB survey of 113,043 adults across 31 Indian states and union territories, published in 2023, found a weighted hypertension prevalence of 35.5%, alongside diabetes at 11.4%, prediabetes at 15.3%, generalised obesity at 28.6%, abdominal obesity at 39.5% and dyslipidaemia at 81.2%. Much of the hypertension was undiagnosed. Two things follow, and the second is the point of this page. First, a reading in the hypertensive range is not unusual and is not a verdict on anybody. Second, the gap between knowing your numbers and not knowing them is where the avoidable harm sits, and that gap is closed by measuring properly and going to a doctor with the readings — not by a classification on a screen.

Related calculators

References

Whelton PK, Carey RM, Aronow WS, et al. Guideline for the Prevention, Detection, Evaluation, and Management of High Blood Pressure in Adults (the 2017 ACC/AHA guideline, issued jointly with nine further professional societies). Hypertension 2018;71:e13–e115; DOI 10.1161/HYP.0000000000000065. Source of the four-band classification used in the AHA/ACC column (normal under 120 and under 80; elevated 120–129 and under 80; stage 1 130–139 or 80–89; stage 2 at or above 140 or at or above 90), of the instruction that a person whose systolic and diastolic fall in two categories is assigned to the higher category, and of the diagnostic requirement for an average of at least two readings obtained on at least two occasions with out-of-office measurement recommended for confirmation. Cited as figures and by that one instruction; no table or passage of the guideline is reproduced.
Jones DW, et al. Guideline for the Prevention, Detection, Evaluation and Management of High Blood Pressure in Adults (the 2025 AHA/ACC guideline, issued jointly with eleven further professional societies). Hypertension and Circulation, published online 14 August 2025; DOI 10.1161/HYP.0000000000000249 and 10.1161/CIR.0000000000001356. VERIFIED on the guideline’s own summary: the four blood pressure categories and their thresholds are UNCHANGED from 2017 — normal under 120 and under 80, elevated 120–129 and under 80, stage 1 130–139 or 80–89, stage 2 at or above 140 or at or above 90. Also the source of the wording that severe hypertension in non-pregnant individuals is defined as blood pressure above 180/120 mmHg. This page cites the 2025 guideline as the current American document and the 2017 one as the origin of the 130/80 threshold.
McEvoy JW, et al. 2024 ESC Guidelines for the management of elevated blood pressure and hypertension. Eur Heart J 2024;45(38):3912–4018; DOI 10.1093/eurheartj/ehae178. Source of the three-category classification used in the ESC column: non-elevated blood pressure as office systolic under 120 mmHg AND diastolic under 70 mmHg; elevated blood pressure as office systolic 120–139 mmHg OR diastolic 70–89 mmHg; hypertension as office systolic at or above 140 mmHg OR diastolic at or above 90 mmHg. Also the source of the guideline’s preference for out-of-office measurement to confirm elevated blood pressure or hypertension, and of its separate flagging of readings at or above 180/110 mmHg for immediate treatment. Cited by figure.
Mancia G, Kreutz R, Brunström M, et al. 2023 ESH Guidelines for the management of arterial hypertension. J Hypertens 2023;41:1874–2071; DOI 10.1097/HJH.0000000000003480. Source of the six-grade classification used in the ESH column (optimal under 120 and under 80; normal 120–129 and 80–84; high-normal 130–139 and/or 85–89; grade 1 140–159 and/or 90–99; grade 2 160–179 and/or 100–109; grade 3 at or above 180 and/or at or above 110), of the isolated systolic and isolated diastolic patterns, of the statement that the category is defined by the highest level of blood pressure whether systolic or diastolic, and of the requirement for at least two separate office visits within four weeks before diagnosing hypertension unless grade 3 hypertension, organ damage or established cardiovascular disease is present. Cited by figure.
Centers for Disease Control and Prevention, About High Blood Pressure and Measure Your Blood Pressure. A United States Government work and therefore usable on a commercial site. VERIFIED: the CDC page gives normal as systolic under 120 and diastolic under 80, elevated as systolic 120–129 with diastolic under 80, and high blood pressure as systolic 130 or higher or diastolic 80 or higher, attributing the definitions to the ACC and the AHA — so the 130/80 threshold itself has a clean Government route. VERIFIED ALSO: the CDC page does NOT carry the stage 1 / stage 2 split, nor any hypertensive-crisis threshold, so those are cited to the guidelines. The measurement instructions on this page are the CDC’s: at least five minutes seated with the back supported, both feet flat and legs uncrossed, nothing to eat or drink for 30 minutes beforehand, bladder emptied, arm resting on a table at chest height, cuff against bare skin and snug but not tight, no talking during the reading, and at least two readings one or two minutes apart.
American Heart Association, Understanding Blood Pressure Readings. Cited for the public-facing statement of the severe-hypertension and hypertensive-emergency threshold as HIGHER THAN 180 systolic and/or HIGHER THAN 120 diastolic, for the symptom list that distinguishes a hypertensive emergency (chest pain, shortness of breath, back pain, numbness, weakness, change in vision, difficulty speaking) and for the instruction to call emergency services when a reading above 180/120 is accompanied by those symptoms. VERIFIED, and worth recording because it has changed: the current advice for a reading above 180/120 WITHOUT symptoms is to wait ONE minute and measure again, not the five minutes that earlier versions of this advice gave.
Ishigami J, Charleston J, Miller ER 3rd, et al. Effects of Cuff Size on the Accuracy of Blood Pressure Readings: The Cuff(SZ) Randomized Crossover Trial. JAMA Intern Med, published online 7 August 2023. 195 adults. Using a regular-size cuff instead of the cuff appropriate to the participant’s arm changed mean systolic blood pressure by —3.6 mmHg in those requiring a small cuff, +4.8 mmHg in those requiring a large cuff and +19.5 mmHg in those requiring an extra-large cuff (all P below 0.001). Cited by figure, and the single most practically useful number on this page for anyone using a home monitor.
Anjana RM, Unnikrishnan R, Deepa M, et al. Metabolic non-communicable disease health report of India: the ICMR-INDIAB national cross-sectional study. Lancet Diabetes Endocrinol 2023. 113,043 participants (33,537 urban, 79,506 rural) across 31 states and union territories, surveyed 2008–2020. Weighted prevalence: hypertension 35.5%, diabetes 11.4%, prediabetes 15.3%, generalised obesity 28.6%, abdominal obesity 39.5%, dyslipidaemia 81.2%. Cited on this page purely as context for how common a reading in the hypertensive range is among this site’s largest readership, and not as a reason to interpret any reading differently.
LICENSING POSITIONS, recorded because they determined what this page could say. (1) The two American guidelines and the two European ones are the work of private professional bodies. Their numeric thresholds are facts and are used as such; their classification tables, criteria lists and prose are not reproduced anywhere on this page, and the comparison table above is a comparison assembled here, which no guideline publishes. One sentence of instruction is quoted from each family because the worse-of-the-two rule cannot be stated more briefly than its authors stated it. (2) No guidance from the UK national institute is used anywhere on this site, including its hypertension diagnostic pathway and its 135/85 home-monitoring threshold, which would have been directly relevant here: its open content licence is United Kingdom-only and forbids displaying the licensed information next to advertising, and this site carries advertising. (3) No material from the international health body whose publications are licensed on NonCommercial terms is used, directly or through a third party that has reprinted it, because copyright does not launder. Its global blood pressure estimates and its HEARTS technical package were therefore not drawn on.
DERIVATIONS AND VERIFICATION PERFORMED FOR THIS PAGE, recorded so they can be checked. (1) Each of the three classifications is implemented as max(step from systolic, step from diastolic) within that scheme, which is what both guideline families instruct. (2) Sweeping every whole-mmHg pair the page accepts (systolic 50 to 300, diastolic 20 to 200) that satisfies systolic at least 5 mmHg above diastolic — 33,341 pairs — yields exactly NINE distinct combinations of the three verdicts plus the care-today flag, and they are totally ordered, each component being non-decreasing across the nine. All nine are reachable. This was checked by running every pair through this page’s own engine and comparing with an independently written implementation of the three classifications; the two agree on all 33,341 pairs. (3) Every category boundary was swept from both sides at 0.01 and 0.1 mmHg, and the pulse-pressure guard was tested with decimal inputs whose difference lands on 5 by binary floating point (130.3 — 125.3 evaluates to 5.000000000000014 and is accepted; 124.99 — 120 is refused). No boundary on this page is reached through a division, so the floating-point categorisation defect that affects BMI-derived categories does not arise here; that was verified rather than assumed. (4) Every branch was tested with an absent, zero, negative, out-of-range and non-numeric input on each of the three fields: all refuse and show the incomplete message, and none falls through the chained comparisons into step 1. (5) The approximation (S + 2D)/3 for mean arterial pressure assumes a diastolic interval roughly twice the systolic one and therefore a heart rate in the usual resting range; it is printed as an approximation and is not used to classify anything.
A note on what a blood pressure category is. It is a label placed on a point of a continuous curve by a committee deciding what a word should be used for. It is not a measurement, it is not a property of a person, and the fact that three committees drew it in three different places is the clearest available demonstration of that. The reading itself is a real measurement, taken with real error, of a quantity that varies by tens of mmHg across a day. Nothing on this page should be read as a statement about anybody. It reports where one number pair falls on three published classifications, and it flags the one reading that should change what somebody does today.

CalcEngines health calculators are for education and for checking arithmetic that has already been decided elsewhere. They are not medical advice, they do not decide what to give, and they do not replace the judgement of a doctor, nurse, midwife or dietitian who knows the person in front of them. Every figure depends on the values you enter and on the assumptions stated on the page — check it against the prescription, the product label and your local policy before acting on it.