Paediatric Weight Estimation Calculator (kg, Age-Based)
Paediatric Weight Estimation Calculator: Six Published Age-Based Formulas and the Spread Between Them
What six published age-based formulas give for the same child — the traditional APLS (age + 4) × 2, the three-band APLS set in use since 2011, Luscombe and Owens’ 3 × age + 7, Best Guess, Nelson’s and Argall’s — with the gap between the highest and the lowest shown as a headline figure rather than a footnote, and the published accuracy on the page: in a validation of 410 children, these formulas landed within 10% of the measured weight between 34% and 45.4% of the time, against a proposed minimum of 70%. A measured weight always replaces an estimate.
An estimated weight is what you use when a child cannot be weighed. It is an estimate from age or length, it carries a wide error in both directions, and a measured weight always replaces it. Check anything calculated from this figure against the measured weight as soon as one is available.
Six published formulas for the same child, and how far apart they are
a child of 4 years 0 months, headline from the 2011 APLS three-band set
The six formulas, their age bands, and where they step
- a
- age in years as a decimal, computed here as m/12 from completed years and months. So 4 years 6 months is a = 4.5, not 4
- m
- age in completed months, which is what selects the band in the three banded formulas. 12×years + months
- W
- an ESTIMATED weight in kilograms. It is not a measurement and nothing on this page turns it into a dose, a volume or a rate
- the 6-year step
- the 2011 APLS set changes equation at the sixth birthday and the two equations do not meet: 2a + 8 at a = 5.917 is 19.83 kg and 3a + 7 at a = 6 is 25.00 kg, a jump of 5.17 kg or 26%. It is the largest discontinuity in the six and it is in the published band set, not in this page
- the 1-year steps
- Best Guess steps from 10.0 to 12.0 kg at the first birthday, a 20% jump. The 2011 APLS set and Nelson’s are both continuous there, because 0.5×12 + 4 and (12 + 9)/2 happen to meet their respective next bands — 10.0 and 10.0 for APLS, 10.0 and 10.0 for Nelson’s
- the slopes
- above seven years the six formulas add 2, 3, 3, 4, 3.5 and 3 kg per year of age respectively, in the order traditional APLS, APLS 2011, Luscombe, Best Guess, Nelson’s, Argall’s. Below seven, Nelson’s 1-to-6 band has a slope of 2 like the traditional formula. Different slopes are why the spread widens with age from 13.7% at ten months to 55.6% at fourteen years, and why no amount of care choosing between them makes an age-based estimate narrow
Worked example
a child of 4 years 0 months, headline from the 2011 APLS three-band set
All six, for the same child. Age 4 years 0 months is m = 48 completed months, a = 4.0 years. Traditional APLS: 2(4 + 4) = 16.0 kg. APLS 2011, in its 1-to-5 band: 2(4) + 8 = 16.0 kg. Luscombe and Owens: 3(4) + 7 = 19.0 kg. Best Guess, in its 1-to-5 band: 2(4 + 5) = 18.0 kg. Nelson's, in its 1-to-6 band: 2(4) + 8 = 16.0 kg. Argall's: 3(4 + 2) = 18.0 kg. The headline is the 2011 set, so 16.0 kg.
The spread, which is the point. Lowest 16.0 kg, highest 19.0 kg, so the spread is 3.0 kg and that is 18.75% of the lowest. Anything calculated from 19.0 kg would be 18.75% larger than the same thing calculated from 16.0 kg. For a four-year-old this is near the narrow end of what these formulas do: the spread is 14.3% at three years, 22.2% at five, 25.0% at six, 42.9% at ten and 55.6% at fourteen, and its minimum anywhere in the range is 13.7%, at ten months old.
What the published accuracy says about that 16.0 kg. In 410 Australian children aged 1 to 11, Luscombe's formula put 45.4% of estimates within 10% of the measured weight, Best Guess 42% and the traditional APLS formula 34%, with mean biases of +0.66 kg, +0.7 kg and −4.2 kg. A systematic review of 98 studies and 1,054,673 children proposes that an acceptable system should achieve 70% within 10% and 95% within 20%; the pooled figure for age-based formulas was 65% within 20%, against that 95% target. So the right way to read 16.0 kg is as the centre of a wide interval, not as a measurement, and the published numbers say roughly that fewer than half of such estimates land within a tenth of the truth.
Watch the sixth birthday. Take the same child two years later. At 5 years 11 months the 2011 set is still in its 1-to-5 band: 2(5.917) + 8 = 19.83 kg. One month later, at 6 years 0 months, it switches: 3(6) + 7 = 25.00 kg. The estimate rises by 5.17 kg, 26%, in a month, and no child does that. The discontinuity is in the published band set and this page flags it rather than letting the reader walk into it. Best Guess has a smaller one at the first birthday, 10.0 kg to 12.0 kg. Luscombe's, Argall's and the traditional APLS formula are continuous because each is a single equation; the 2011 set and Nelson's are continuous at one year because their bands happen to meet there exactly.
Age 0, and the top of the range. At birth the six give 8.0, 4.0, 7.0, 4.5, 4.5 and 6.0 kg — a spread of 100% of the lowest, for a baby who typically weighs about 3.5 kg. The traditional APLS formula's 8.0 kg is more than double a term birth weight. The page computes under a month and prints a loud note, because the honest answer for a neonate is the recorded birth weight. At the other end it refuses above 14 years: the widest-stated formula on the page is stated only to 14, and through puberty age tells you almost nothing about mass — at twelve the six formulas already span 32 to 48 kg.
The direction of the error depends on where the child is. This is the finding that matters most for readers in South Asia and it is in the opposite direction from the one the UK literature taught. In the UK, the traditional formula underestimated measured weight by 18.8% in 17,244 children, which is why Luscombe's replacement exists. In India, an analysis of 171,738 children aged 7 to 59 months found every age-based formula tested OVERestimating: the traditional APLS formula by a mean of 3.51 kg, Luscombe's by 4.09 kg and Argall's by 6.50 kg. A study of 500 children in Bangalore found the same direction: the traditional formula over by 2.23 kg, Argall's by 2.38 kg. Overestimation is the hazardous direction. The page prints illustrative rows under five years showing what subtracting those published mean biases would leave — labelled as illustrations, because nobody has published a corrected formula and subtracting a population mean from an individual is not a method.
What to do instead, where you can. Measure. Failing that, measure a length and use a tape, or ask the parent. In the same systematic review, a length-based tape achieved about 56% of estimates within 10% of measured weight, a habitus-adjusted tape about 78% and a method using mid-arm circumference and humeral length about 71%, while a parent's estimate of their own child achieved about 70% — better than every formula on this page. The tapes are copyrighted products and are not reproduced here; the point of naming them is that they exist and they are better. One caution specific to India: even a length-based tape overestimates there, by 5% to 15% depending on the band in a Chennai study of 1,185 children, which is why an India-adjusted version of it was developed.
What this page will not do. No dose, no drug, no infusion volume, no defibrillation energy, no tube size. It outputs a weight and stops, which is what keeps it in scope. It is also not the page for a premature baby's corrected age, which is the corrected age page's job, nor for postnatal weight loss, which is the newborn weight loss page's. Paediatric maintenance fluid from a known weight is on the paediatric maintenance fluid page.
All six formulas across the paediatric age range, and the spread between them
| Age | Traditional APLS | APLS 2011 | Luscombe | Best Guess | Nelson’s | Argall’s | Lowest to highest | Spread |
|---|---|---|---|---|---|---|---|---|
| Birth | 8.00 | 4.00 | 7.00 | 4.50 | 4.50 | 6.00 | 4.00 to 8.00 | 4.00 kg (100.0%) |
| 1 month | 8.17 | 4.50 | 7.25 | 5.00 | 5.00 | 6.25 | 4.50 to 8.17 | 3.67 kg (81.5%) |
| 6 months | 9.00 | 7.00 | 8.50 | 7.50 | 7.50 | 7.50 | 7.00 to 9.00 | 2.00 kg (28.6%) |
| 10 months | 9.67 | 9.00 | 9.50 | 9.50 | 9.50 | 8.50 | 8.50 to 9.67 | 1.17 kg (13.7%) |
| 1 year | 10.00 | 10.00 | 10.00 | 12.00 | 10.00 | 9.00 | 9.00 to 12.00 | 3.00 kg (33.3%) |
| 2 years | 12.00 | 12.00 | 13.00 | 14.00 | 12.00 | 12.00 | 12.00 to 14.00 | 2.00 kg (16.7%) |
| 3 years | 14.00 | 14.00 | 16.00 | 16.00 | 14.00 | 15.00 | 14.00 to 16.00 | 2.00 kg (14.3%) |
| 4 years | 16.00 | 16.00 | 19.00 | 18.00 | 16.00 | 18.00 | 16.00 to 19.00 | 3.00 kg (18.8%) |
| 5 years 11 months | 19.83 | 19.83 | 24.75 | 23.67 | 19.83 | 23.75 | 19.83 to 24.75 | 4.92 kg (24.8%) |
| 6 years | 20.00 | 25.00 | 25.00 | 24.00 | 20.00 | 24.00 | 20.00 to 25.00 | 5.00 kg (25.0%) |
| 8 years | 24.00 | 31.00 | 31.00 | 32.00 | 25.50 | 30.00 | 24.00 to 32.00 | 8.00 kg (33.3%) |
| 10 years | 28.00 | 37.00 | 37.00 | 40.00 | 32.50 | 36.00 | 28.00 to 40.00 | 12.00 kg (42.9%) |
| 12 years | 32.00 | 43.00 | 43.00 | 48.00 | 39.50 | 42.00 | 32.00 to 48.00 | 16.00 kg (50.0%) |
| 14 years | 36.00 | 49.00 | 49.00 | 56.00 | 46.50 | 48.00 | 36.00 to 56.00 | 20.00 kg (55.6%) |
Published accuracy: what proportion of estimates land within 10% of the measured weight
| Method | Within 10% of measured weight | Within 20% | Mean bias | Population |
|---|---|---|---|---|
| Habitus-adjusted length tape | 78.0% | 96.6% | — | Pooled, 98 studies |
| Mid-arm circumference and humeral length method | 70.9% | 95.3% | — | Pooled, 98 studies |
| A parent’s estimate of their own child | 69.8% | 87.1% | — | Pooled, 98 studies |
| Length-based tape | 55.6% | 81.2% | — | Pooled, 98 studies |
| Age-based formulas, pooled | not reported in the open text | 65.0% | — | Pooled, 98 studies |
| Luscombe, (3 x age) + 7 | 45.4% | not reported | +0.66 kg | 410 children, 1–11 y, Australia |
| Best Guess | 42% | not reported | +0.7 kg | 410 children, 1–11 y, Australia |
| Traditional APLS, (age + 4) x 2 | 34% | not reported | −4.2 kg | 410 children, 1–11 y, Australia |
| PROPOSED MINIMUM for an acceptable system | above 70% | above 95% | — | — |
Direction of error by population, which reverses between the UK and India
| Study | Population | Formula | Finding |
|---|---|---|---|
| Luscombe & Owens 2007 | 17,244 children, 1–10 y, large UK emergency department | Traditional APLS | UNDERestimated by a mean of 18.8% (95% CI 18.42 to 19.18) |
| Luscombe & Owens 2007 | same | (3 x age) + 7 | UNDERestimated by a mean of 2.48% (95% CI 2.17 to 2.79) on its own derivation sample |
| Kelly, Nguyen & Krieser 2011 | 410 children, 1–11 y, Australian metropolitan ED | Luscombe / Best Guess / traditional APLS | Mean bias +0.66 / +0.7 / −4.2 kg; within 10% of measured 45.4% / 42% / 34% |
| Seddon et al. 2012 | 599 children, multiethnic UK population, 157 Asian | APLS 2011 three-band set | Infants: bias 0.27 kg, not significant. 1–5 y: UNDERestimated by 1.3 kg (9.4%), p below 0.001. 6–10 y: 0.42% under. Asian 1–5 y subgroup 1.0 kg under against 1.3 kg in Caucasian children |
| Varghese et al. 2006 | 500 children, 0.1–11.4 y, Bangalore, most under 15 kg | Traditional APLS / Argall / Nelson 1–6 y / Nelson 3–12 mo | OVERestimated by 2.23 (SD 1.69) / 2.38 (SD 2.73) / 2.29 (SD 1.76) / 0.69 (SD 1.26) kg. Length-based tape bias 0.080 kg (SD 0.96) |
| Sharma et al. 2023 | 171,738 children, 7–59 months, Indian national family health survey | Traditional APLS / Luscombe / Argall | All three OVERestimated. Mean difference 3.51 kg (SD 2.08) for APLS, which was the smallest, 4.09 kg for Luscombe and 6.50 kg for Argall |
| Asskaryar & Shankar 2015 | 1,185 children, 1 month to 12 years, Chennai | Length-based tape | OVERestimated by 5% to 15% depending on band; correct band in 33% to 86.6% of children. An India-adjusted tape applying an 8% correction raised that to 51% to 97.8% |
Where each figure on this page comes from, and what was deliberately left out
| Figure | Source used | What was rejected, or could not be established |
|---|---|---|
| Traditional APLS, (age + 4) x 2 | Its equation needs no source: it is arithmetic, quoted identically in every text that mentions it, and it is reproduced here as the thing those texts quote. Its accuracy is taken from Luscombe M, Owens B, Arch Dis Child 2007;92(5):412–15 and from Kelly A-M, Nguyen K, Krieser D, Emerg Med Australas 2011;23:59–62. | An origin. The formula’s provenance is not clear from the available literature; it is variously attributed to mid-twentieth-century paediatric texts, and this page does not assert a first source it cannot verify. |
| APLS 2011 three-band set | Stated as the three equations validated in Seddon C, Lockitt L, Dhanjal S, Eisenhut M, Validation of Advanced Paediatric Life Support formulas for weight calculation in a multiethnic population, ISRN Pediatrics 2012;2012:869634 — (0.5 x age in months) + 4 for 1 to 11 months, (2 x age) + 8 for 1 to 5 years, (3 x age) + 7 for 6 to 12 years. Cited to the peer-reviewed validation rather than to the course manual. | The course manual itself, which is copyrighted and is not reproduced, quoted or paraphrased anywhere on this page. Only the equations — which are arithmetic — and the validation’s published results are used. |
| Luscombe and Owens, (3 x age) + 7 | Luscombe M, Owens B. Weight estimation in resuscitation: is the current formula still valid? Arch Dis Child 2007;92(5):412–15. 17,244 children aged 1 to 10 attending a large emergency department in a major UK city, June to December 2005. | A city. The paper describes its setting as a large ED in a major UK city without naming it, and its two authors are affiliated to hospitals in two different cities, so this page does not name one. Several secondary sources do, and they disagree with each other. |
| Best Guess | Tinning K, Acworth J, Make your Best Guess: an updated method for paediatric weight estimation in emergencies, Emerg Med Australas 2007, derived on 70,181 children at a Brisbane children’s hospital. External validation: Kelly A-M, Kerr D, Clooney M, Krieser D, Nguyen K, Emerg Med Australas 2007;19(6):543–6. | Certainty about where its middle band ends. Sources state it as 1 to 5 years and the next band as either 5 to 14 or 6 to 14. It makes no difference: 2(a + 5) and 4a are exactly equal at a = 5, so the two expressions meet without a step and the boundary can be put at either. This page puts it at 5 years and says why. |
| Nelson’s formula set | Taken as stated in Varghese A, Vasudevan VK, Lewin S, Indumathi CK, Dinakar C, Rao SD, Do the length-based (Broselow) tape, APLS, Argall and Nelson’s formulae accurately estimate weight of Indian children?, Indian Pediatr 2006;43:889–94, which gives the three bands as (months + 9)/2 for 3 to 12 months, (age x 2) + 8 for 1 to 6 years and (age x 7 − 5)/2 for 7 to 12 years. | The textbook it is named for, which is copyrighted and not consulted or reproduced. Note that its infant band is identical to Best Guess’s and its 1-to-6 band is identical to the 2011 APLS 1-to-5 band, which is why three of the six columns coincide at many ages. |
| Argall’s, (age + 2) x 3 | Argall JA, Wright N, Mackway-Jones K, Jackson R, A comparison of two commonly used methods of weight estimation, Arch Dis Child 2003;88:789–90, as cited in the Indian validations above, which report its bias directly. | A within-10% accuracy figure for this formula from a primary source. One is widely quoted (37%) but could not be traced to the paper in the material available here, so it is not stated on this page. |
| Accuracy benchmark and pooled figures | Wells M, Goldstein LN, Bentley A. The accuracy of emergency weight estimation systems in children — a systematic review and meta-analysis. Int J Emerg Med 2017;10:29. 98 studies, 1,054,673 patients. Source of the proposed benchmark (above 70% within 10% and above 95% within 20%) and of the pooled figures for the habitus tape (78.0% / 96.6%), the mid-arm method (70.9% / 95.3%), parental estimate (69.8% / 87.1%), the length tape (55.6% / 81.2%) and age-based formulas (65.0% within 20%). | The pooled WITHIN-10% figure for age-based formulas. It could not be extracted from the openly available text, which reports the pooled within-20% figure and describes age-based accuracy as very low without giving the within-10% number in the text available here. It is left as not established rather than guessed. |
| South Asian validation data | Varghese et al. 2006 (500 children, Bangalore); Sharma S, Feroz SH, Yadav J, Rao MVV, Validation of three age based weight formulae for estimating weight among Indian children, Int J Community Med Public Health 2023;10(4):1517–20 (171,738 children aged 7 to 59 months from a national family health survey); Seddon et al. 2012 for the Asian subgroup of a multiethnic UK cohort. Used for the direction and size of error, and for the illustrative rows under five years. | A weight-estimation formula DERIVED in a South Asian population. None was found. The 2023 analysis concludes that the formulas need adjusting for Indian children, which is a statement that the adjustment does not yet exist. This page therefore does not offer a South Asian formula and does not invent one: it shows the published mean biases, labels the subtraction as an illustration, and says plainly that subtracting a population mean from an individual is not a method. Also noted and not used: a secondary summary of a newer multicentre Indian tape study claiming 78.6% of estimates within 10% in 2,253 children across 13 intensive care units, which could not be traced to a primary citation and is therefore not quoted here. |
| Length- and habitus-based tapes | Named, and their published accuracy quoted from the systematic review above and from Asskaryar F, Shankar R, An Indian pediatric emergency weight estimation tool: prospective adjustment of the Broselow tape, Int J Emerg Med 2015;8:28 (1,185 children, Chennai). | The tapes themselves. Both the length-based tape and the habitus-adjusted tape are copyrighted products with their own band tables, and no part of either is reproduced, redrawn or approximated here. The page says they exist, says they are more accurate, and says to use one where it is available. Reproducing a band table would also be the one thing on this page that could be used as a dosing aid, which is outside its scope by design. |
| Anything expressed as a dose | Nothing. The page has no drug, no dose, no concentration and no volume anywhere in it, by design. | Every form of dosing illustration that names a drug or a quantity. Where the page needs to show why the spread matters, it says that a quantity calculated from the top of the range exceeds the same quantity calculated from the bottom by the stated percentage, and names nothing. |
Why this page prints six numbers and the gap between them instead of one confident estimate
An estimated weight exists for one situation: a child who cannot be weighed, now. A child in cardiac arrest, a child on a resuscitation trolley, a child being moved. Everything downstream — drug doses, fluid volumes, energies, tube sizes — is then calculated from a number that is not a measurement, and the whole of the uncertainty in those calculations comes from this one step. That is why this page is built the way it is: it prints six published formulas side by side, and it prints the gap between the highest and the lowest as a headline figure rather than a footnote, because that gap is the uncertainty and a reader who sees one number has no way to know it exists. The moment a measured weight is available, it replaces the estimate and everything calculated from the estimate is recalculated. That is not a disclaimer; it is the only correct use of this page.
The formulas disagree, and they disagree more the older the child. They are not variants of one method. They were derived on different populations in different decades with different body sizes, and above seven years of age they add 2, 3, 3, 4, 3.5 and 3 kilograms per year of age respectively, in the order traditional APLS, APLS 2011, Luscombe, Best Guess, Nelson’s, Argall’s. Different slopes compound: the six span 1.17 kg at ten months old, which is their closest agreement anywhere and still 13.7% of the lowest; 3.0 kg at four years, which is 18.8%; 12 kg at ten years, which is 42.9%; and 20 kg at fourteen, which is 55.6%. The traditional (age + 4) × 2 formula is the lowest of the six at every age from two years upward, joint-lowest with two others through early childhood, and lowest outright from about seven years by a widening margin — at ten years it gives 28 kg where Best Guess gives 40. There is no version of choosing carefully between them that makes an age-based estimate narrow, because the width is not noise in any one of them, it is the fact that age does not determine mass.
Two of the published band sets step discontinuously, which is a property of the sources and not of this page. The three-band APLS set changes equation at the sixth birthday, and the two equations do not meet: at 5 years 11 months it gives 2(5.917) + 8 = 19.83 kg, and at 6 years 0 months it gives 3(6) + 7 = 25.00 kg. That is a rise of 5.17 kg, 26% of the lower figure, in one month of age, and no child does that. Best Guess has a smaller version at the first birthday, 10.0 kg to 12.0 kg, a 20% step. The other four are continuous — Luscombe’s, Argall’s and the traditional APLS formula because each is one equation for the whole range, and Nelson’s because its infant band happens to meet its next band exactly at twelve months. This page flags a child who is within a few months of the sixth birthday, because at that point the two figures either side of the step are both defensible and the honest reading is the interval between them rather than whichever one the calendar selects.
The published accuracy is on this page because it is the number a clinician needs and almost nothing shows it. In 410 Australian children aged 1 to 11, the proportion of estimates landing within 10% of the measured weight was 45.4% for Luscombe’s formula, 42% for Best Guess and 34% for the traditional APLS formula, with mean biases of +0.66 kg, +0.7 kg and −4.2 kg. A systematic review of 98 studies and 1,054,673 children proposes that an acceptable weight-estimation system should put 70% of estimates within 10% of the measured weight and 95% within 20%; no age-based formula is anywhere near that, and the pooled age-based figure for the easier 20% criterion is 65% against a 95% target. Everything that uses a measurement does better. A length-based tape achieves about 56% within 10%, a habitus-adjusted tape about 78%, a method using mid-arm circumference and humeral length about 71%, and a parent asked to estimate their own child’s weight about 70% — better than every formula on this page. If a scale, a tape or a parent is available, use it. The tapes are copyrighted products and nothing of them is reproduced here; naming them is the point, because they are better and a reader should know that before using an age formula.
The direction of the error reverses between populations, and that matters more than its size. The reason Luscombe’s formula was written in 2007 is that the traditional formula had fallen behind the British child: in 17,244 UK children aged 1 to 10 it underestimated measured weight by a mean of 18.8%. Underestimating is the safer direction for most calculations. In Indian children the picture inverts. An analysis of 171,738 children aged 7 to 59 months from a national family health survey found every age-based formula tested OVERestimating: the traditional APLS formula by a mean of 3.51 kg, which was the smallest of the three, Luscombe’s by 4.09 kg and Argall’s by 6.50 kg. A study of 500 children in Bangalore found the same direction, the traditional formula over by 2.23 kg and Argall’s by 2.38 kg. The systematic review reaches the general version of this: studies in high-income countries have mostly shown underestimation, while studies in low- and middle-income countries have mostly shown overestimation, potentially to a dangerous degree. Even a length-based tape overestimates in Indian children, by 5% to 15% depending on the band in a Chennai study of 1,185 children, which is why an India-adjusted version of it was developed. For readers under five this page prints illustrative rows showing what subtracting the published mean biases would leave, and labels them as illustrations — because no corrected formula has been published, subtracting a population mean from an individual is not a method, and the standard deviations around those means are large.
Age 0 and the top of the range, both decided deliberately. At birth the six formulas give 8.0, 4.0, 7.0, 4.5, 4.5 and 6.0 kg for a baby who typically weighs about 3.5 kg — a spread of 100% of the lowest, and the traditional formula’s 8.0 kg is more than double a term birth weight. The page computes under a month rather than refusing, because an unweighed three-week-old in an emergency is a real situation, but it prints a loud note saying that a newborn’s weight is recorded at birth and that the record is the figure to use. At the other end the page refuses above fourteen years. Fourteen is the upper limit of the widest-stated formula on it, and beyond that the objection is not bookkeeping: puberty happens at widely different ages, two thirteen-year-olds can differ by thirty kilograms, and no function of age alone can know which is in front of you. The spread shows it — at twelve years the six formulas already span 32 to 48 kg.
What this page does not do, and where those questions live. It outputs a weight and nothing else: no dose, no drug, no infusion volume, no energy, no tube size. That boundary is what keeps it in scope, and it is why the page describes the consequence of the spread as a percentage rather than as a quantity of anything. Correction for prematurity is the corrected age page’s subject and is not duplicated here; postnatal weight loss belongs to the newborn weight loss page; maintenance fluid from a known weight is the paediatric maintenance fluid page, and glucose infusion rate its own page. Growth and BMI centiles are not here either, and that is a licensing matter rather than a choice: the reference data that varies continuously with age comes from bodies whose licences this site cannot use.
Frequently asked questions
Which of the six formulas should I use?
Whichever your local resuscitation policy names, so that your figure is the same as your colleagues’. Beyond that, read the spread rather than choosing. The evidence that exists ranks them loosely: in 410 Australian children aged 1 to 11, Luscombe’s (3 x age) + 7 put 45.4% of estimates within 10% of measured weight, Best Guess 42% and the traditional (age + 4) x 2 only 34%, so the traditional formula is the one with the clearest case against it in a Western population. But in Indian children the ranking inverts: in a national-survey analysis of 171,738 children under five, the traditional formula had the SMALLEST mean error of the three tested and Luscombe’s the largest, because all three overestimated and Luscombe’s overestimates most. So there is no answer that holds everywhere, which is exactly why the page prints all six and the gap between them.
Why does the estimate jump when the child turns six?
Because the three-band APLS set changes equation at the sixth birthday and the two equations do not meet. At 5 years 11 months it uses (2 x age) + 8 and gives 19.83 kg; at 6 years 0 months it uses (3 x age) + 7 and gives 25.00 kg. That is 5.17 kg, or 26%, in one month, and it is a property of the published band set, not an error in this page. No child gains 5 kg in a month. If your patient is near that birthday, the honest reading is the interval between the two figures, and the page flags it when the age is within a few months either side. Best Guess has a smaller version at the first birthday, stepping from 10.0 to 12.0 kg. The single-equation formulas — the traditional APLS one, Luscombe’s and Argall’s — have no steps at all, which is the one thing they have going for them.
How accurate are these formulas, really?
Poorly, and the page prints the numbers rather than saying “approximately”. The benchmark proposed in a systematic review of 98 studies and 1,054,673 children is that an acceptable system should put 70% of estimates within 10% of the measured weight and 95% within 20%. Against that, the three age-based formulas directly compared in a 410-child validation achieved 45.4%, 42% and 34% within 10%, and the pooled age-based figure for the easier 20% criterion was 65% against a 95% target. The review’s own summary is that age-based estimates achieve very low accuracy. For comparison, the same review’s pooled figures were 55.6% within 10% for a length-based tape, 78.0% for a habitus-adjusted tape, 70.9% for a method using mid-arm circumference and humeral length, and 69.8% for a parent asked to estimate their own child’s weight. One figure is deliberately missing from the page: the pooled within-10% accuracy of age-based formulas, which could not be extracted from the openly available text of the review and is marked as not established rather than guessed at.
Should I use a Broselow or PAWPER tape instead?
If one is available, yes, and the published figures say so clearly: a length-based tape puts about 56% of estimates within 10% of measured weight against 34% to 45.4% for the formulas here, and a habitus-adjusted tape about 78%, which is the only method in the review that clears the proposed 70% benchmark. Neither tape is reproduced on this page, in any form: they are copyrighted products with their own band tables, and redrawing or approximating one would also be the one thing here that could be used directly as a dosing aid, which is outside this page’s scope by design. There is one caution for readers in India. In 1,185 children in Chennai, the standard length-based tape overestimated weight by 5% to 15% depending on the band and identified the correct band only 33% to 86.6% of the time, which is why an India-adjusted version applying an 8% correction was developed and reached 51% to 97.8%. A tape is better than a formula there too, but it is not immune to the same population problem.
Why does the page go up to 14 years and no further?
Because fourteen is the top of the stated range of the widest formula on the page, and because beyond it age tells you almost nothing. Puberty arrives at widely different ages, so two thirteen-year-olds of the same birthday can differ by thirty kilograms, and a function whose only input is age cannot know which one is in front of you. You can watch it happen in the spread: at ten years the six formulas span 28 to 40 kg, at twelve 32 to 48, at fourteen 36 to 56. Extending any of them into adolescence would produce a confident-looking number with nothing behind it, so the page refuses rather than extrapolating. Above this age the approach is an adult one: weigh the patient, or measure a length and use a tape.
What should I do for a newborn or a baby a few weeks old?
Use the recorded birth weight, allowing for the normal postnatal loss and regain. A newborn’s weight is written down, which is the one situation where an estimate is never needed. Look at what the formulas give at birth and the reason is obvious: 8.0, 4.0, 7.0, 4.5, 4.5 and 6.0 kg for a baby who typically weighs about 3.5 kg, a spread of 100% of the lowest, and the traditional APLS formula’s figure is more than double a term birth weight. The page does compute below a month, because an unweighed three-week-old in an emergency is a real situation and refusing would not help anybody, but it prints a loud note. Note also that only three of the six formulas have an infant band at all; the other three are being extrapolated below one year. Postnatal weight loss as a percentage is a different calculation and belongs to the newborn weight loss page in the medical set, and corrected age for prematurity belongs to the corrected age page.
The formulas overestimate in Indian children. Can I just subtract the difference?
No, and the page is careful about this. It does print rows for children under five showing what subtracting the published mean biases would leave — 3.51 kg off the traditional APLS estimate and 4.09 kg off Luscombe’s, from an analysis of 171,738 Indian children aged 7 to 59 months, and 2.23 kg off the traditional estimate from a Bangalore cohort of 500 children nearly all under 15 kg — and it labels those rows as illustrative only. Three reasons they are not a formula. Nobody has published a corrected age-based formula for a South Asian population; the 2023 analysis concludes that one is needed, which is a statement that it does not exist. The standard deviations around those means are large relative to the means themselves, 2.08 kg on a 3.51 kg bias in the national-survey study. And the data come from population surveys rather than from children presenting as emergencies, who are not the same group. The rows are there so a reader in India can see how big the problem is, not so that the arithmetic gets done.
Why does this page not calculate the drug dose as well?
Because deciding what to give is not what this site does, on any page, and because a weight estimate and a dose calculation are two different kinds of uncertainty that should not be hidden inside one number. The page outputs a weight and stops. What it will tell you is the consequence of the spread in general terms: anything calculated from the highest of the six estimates exceeds the same thing calculated from the lowest by exactly the percentage in the spread row, which for a ten-year-old is 42.9%. That is the figure to carry into whatever the weight is being used for, and it is the reason to get a measured weight as soon as one can be obtained and recalculate everything from it.
Why do three of the columns often show the same number?
Because three of the six formulas share equations. Nelson’s infant band and Best Guess’s infant band are both (age in months + 9)/2, so those two columns are identical below one year. Nelson’s 1-to-6-year band and the 2011 APLS 1-to-5-year band are both (2 x age) + 8, so those agree through most of early childhood, and the traditional (age + 4) x 2 is algebraically the same expression again — 2a + 8 — which is why the traditional APLS, APLS 2011 and Nelson columns all read alike from one to five years and then separate. That is worth knowing, because it means the apparent agreement of three formulas in that age band is not three independent confirmations; it is one equation written three times. The spread row is computed from the distinct values, so it is not fooled by this, but a reader scanning the column of six might be.
Related calculators
References
- Luscombe M, Owens B. Weight estimation in resuscitation: is the current formula still valid? Arch Dis Child 2007;92(5):412–15. 17,244 children aged 1–10 years attending a large emergency department in a major UK city between June and December 2005. The traditional APLS formula, weight = 2(age + 4), underestimated measured weight by a mean of 18.8% (95% CI 18.42% to 19.18%); the proposed replacement, weight = 3(age) + 7, underestimated by a mean of 2.48% (95% CI 2.17% to 2.79%) on the same sample. NOTE on provenance: the paper gives its setting as a large ED in a major UK city without naming it, and its authors are affiliated to hospitals in two different cities, so this page does not name one — secondary sources that do, disagree.
- Kelly A-M, Nguyen K, Krieser D. Validation of the Luscombe weight formula for estimating children’s weight. Emerg Med Australas 2011;23:59–62. 410 children aged 1–11 years at an Australian metropolitan teaching hospital emergency department; median age 4 years, mean weight 21.2 kg. The source of the within-10% figures printed on this page: Luscombe’s formula 45.4% of estimates within 10% of measured weight with a mean bias of +0.66 kg (root mean square error 5.44 kg, 95% limits of agreement −9.9 to +11.3 kg); Best Guess 42% with a mean bias of +0.7 kg; the traditional APLS formula 34% with a mean bias of −4.2 kg. This is the single most useful accuracy comparison available for these three formulas and it is why the page can state the figures per formula rather than in general terms.
- Wells M, Goldstein LN, Bentley A. The accuracy of emergency weight estimation systems in children — a systematic review and meta-analysis. Int J Emerg Med 2017;10:29. 98 studies in the meta-analysis, 1,054,673 patients. Source of the benchmark this page quotes — that a weight estimation system should demonstrate more than 70% of estimates within 10% of actual weight and more than 95% within 20% — and of the pooled figures: habitus-adjusted tape 78.0% within 10% and 96.6% within 20%; mid-arm circumference and humeral length method 70.9% and 95.3%; parental estimate 69.8% and 87.1%; length-based tape 55.6% and 81.2%; pooled age-based formulas 65.0% within 20%. Also the source of the statement on this page that studies in high-income countries have mostly shown underestimation of weight while studies in low- and middle-income countries have mostly shown overestimation, potentially to a dangerous degree. WHAT COULD NOT BE ESTABLISHED: the pooled WITHIN-10% figure for age-based formulas. It is not in the openly available text consulted here, which gives the within-20% figure and characterises age-based accuracy as very low. The page records it as not established rather than filling it in.
- Seddon C, Lockitt L, Dhanjal S, Eisenhut M. Validation of Advanced Paediatric Life Support formulas for weight calculation in a multiethnic population. ISRN Pediatrics 2012;2012:869634. 599 children prospectively enrolled; 157 Asian, 268 Caucasian, 174 of other origin. Cited for two things. It is the peer-reviewed statement of the 2011 three-band set this page implements — (0.5 x age in months) + 4 for 1 to 11 months, (2 x age in years) + 8 for 1 to 5 years, (3 x age in years) + 7 for 6 to 12 years — which means the equations can be taken from a journal article rather than from the copyrighted course manual. And it is the only validation found that reports an Asian subgroup: in 1-to-5-year-olds the set underestimated by a mean of 1.3 kg (9.4%, p below 0.001) overall, by 1.0 kg in the 98 Asian children and by 1.3 kg in the 114 Caucasian children; in infants the mean bias was 0.27 kg and not statistically significant, at 6 to 10 years the underestimate was 0.42%, and at 11 to 12 years there was no significant difference.
- Varghese A, Vasudevan VK, Lewin S, Indumathi CK, Dinakar C, Rao SD. Do the length-based (Broselow) tape, APLS, Argall and Nelson’s formulae accurately estimate weight of Indian children? Indian Pediatr 2006;43:889–94. 500 children aged 0.1–11.4 years at a tertiary teaching hospital outpatient department in Bangalore; mean weight 8.19 kg, 458 of them under 15 kg. In the under-15 kg group the mean differences were: length-based tape 0.080 kg (SD 0.96), Nelson’s 3-to-12-month band 0.69 kg (SD 1.26), traditional APLS 2.23 kg (SD 1.69), Nelson’s 1-to-6-year band 2.29 kg (SD 1.76) and Argall’s 2.38 kg (SD 2.73), all in the direction of OVERestimation. Correlation with measured weight: tape r = 0.974, APLS r = 0.902, Argall r = 0.902, Nelson combined r = 0.935. This is also the source from which this page takes Nelson’s three bands as equations. The paper reports mean bias and correlation and does NOT report within-10% proportions, which is stated here rather than inferred.
- Sharma S, Feroz SH, Yadav J, Rao MVV. Validation of three age based weight formulae for estimating weight among Indian children. Int J Community Med Public Health 2023;10(4):1517–20. Secondary analysis of 171,738 live births aged 7–59 months in the fourth National Family Health Survey (2015–16), covering 37 states and union territories. All three formulas overestimated: the traditional APLS formula showed the smallest mean weight difference at 3.51 kg (SD 2.08), Luscombe’s 4.09 kg and Argall’s 6.50 kg. The paper concludes that the formulas need adjusting for Indian children. PROVENANCE NOTE: the abstract also gives percentage overestimates of 23%, 38% and 28% for APLS, Luscombe and Argall respectively, which do not rank in the same order as the kilogram figures printed beside them (Luscombe’s 4.09 kg is smaller than Argall’s 6.50 kg, yet 38% is larger than 28%). The discrepancy could not be reconciled from the material available, so only the kilogram figures are used on this page and the percentages are not quoted.
- Asskaryar F, Shankar R. An Indian pediatric emergency weight estimation tool: prospective adjustment of the Broselow tape. Int J Emerg Med 2015;8:28. 1,185 children aged 1 month to 12 years in Chennai, 769 in development and 416 in validation. The standard length-based tape identified the correct weight band in 33% to 86.6% of children depending on the zone, with a mean error of 5% to 15% in the direction of OVERestimation; an India-adjusted version applying an 8% correction factor reached 51% to 97.8%. Cited on this page to make a point that matters locally: in Indian children even the measurement-based method, which outperforms every age formula here, errs in the dangerous direction, so switching from a formula to a tape does not by itself remove the population problem.
- Tinning K, Acworth J. Make your Best Guess: an updated method for paediatric weight estimation in emergencies. Emerg Med Australas 2007, derived on 70,181 children at a Brisbane children’s hospital; external validation in Kelly A-M, Kerr D, Clooney M, Krieser D, Nguyen K, External validation of the Best Guess formulae for paediatric weight estimation, Emerg Med Australas 2007;19(6):543–6, which found a mean bias of 0.9 kg with 95% limits of agreement −3.5 to +5.3 kg and 76% of estimates within 20% in 1-to-5-year-olds, and a mean bias of 0.4 kg with limits of agreement −14.4 to +15.2 kg and 64% within 20% in 5-to-11-year-olds, concluding that the formulas performed moderately well but tended to overestimate, particularly in children with a lower body mass index. NOTE on its bands: sources state the upper band as either 5 to 14 or 6 to 14 years. It makes no numerical difference, because 2(age + 5) and 4 x age are exactly equal at age 5, so the two expressions meet without a step; this page puts the boundary at 5 years.
- Argall JA, Wright N, Mackway-Jones K, Jackson R. A comparison of two commonly used methods of weight estimation. Arch Dis Child 2003;88:789–90 — the source of the (age + 2) x 3 formula this page includes as Argall’s. Its accuracy on this page is taken from the two Indian validations above, which report its bias directly. WHAT COULD NOT BE ESTABLISHED: a within-10% proportion for this formula from a primary source. A figure of 37% is widely quoted in secondary material and could not be traced to the paper in what was available here, so it is not stated on the page.
- LICENSING POSITION taken for this page, recorded because it determined what the page contains. The six equations are arithmetic and are used freely; the course manual in which the APLS set is published is copyrighted and is neither quoted nor paraphrased anywhere, which is why the 2011 equations are cited to a peer-reviewed validation instead. Both the length-based and the habitus-adjusted tapes are copyrighted products: this page names them, states their published accuracy, and reproduces no part of either — no band table, no colour zones, no redrawn or approximated equivalent. That restraint is not only legal. A reproduced band table is the one artefact on a page like this that could be used directly as a dosing aid, and this page outputs a weight and nothing else by design. No reference-growth data is used, so the licensing questions that attach to growth standards do not arise here. Separately, and as elsewhere on this site, no guidance from the UK national institute is used: its open content licence is United Kingdom-only and forbids display of the licensed information next to advertising, and this site carries advertising.
- Derivations and checks performed for this page rather than taken from a source, recorded so they can be checked. (1) The spread between the six formulas, as a percentage of the lowest, has a minimum of 13.7% at ten completed months and rises monotonically above about a year: 16.7% at two years, 14.3% at three, 18.8% at four, 25.0% at six, 33.3% at eight, 42.9% at ten, 50.0% at twelve and 55.6% at fourteen. At birth it is 100%. (2) The only large discontinuity in the six is the 2011 APLS band change at 72 completed months: 2(71/12) + 8 = 19.833 kg against 3(6) + 7 = 25.000 kg, a step of 5.167 kg or 26.1%. Best Guess steps 10.0 to 12.0 kg at 12 months, 20.0%. Every other band junction is continuous: the 2011 set at 12 months (10.0 to 10.0), Nelson’s at 12 months (10.0 to 10.0) and at 84 months (22.0 to 22.0), and Best Guess at 60 months (20.0 to 20.0). (3) The traditional APLS formula, 2(a + 4), expands to 2a + 8, which is algebraically identical to the 2011 set’s 1-to-5 band and to Nelson’s 1-to-6 band — so three of the six columns coincide exactly through early childhood and that apparent agreement is one equation, not three. (4) Above seven years the six formulas have slopes of 2, 3, 3, 4, 3.5 and 3 kg per year of age (traditional APLS, APLS 2011, Luscombe, Best Guess, Nelson’s, Argall’s), which is why the spread widens with age; below seven, Nelson’s 1-to-6 band shares the traditional formula’s slope of 2. (5) At four completed months one month of age error moves the three infant formulas by 0.5 kg, about 8% of the estimate.
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.