CSF Glucose Unit Converter

Q: How do I convert CSF glucose from mg/dL to mmol/L?

Multiply by 0.0555074, the same factor as for plasma glucose, because it is the same molecule with a molar mass of 180.156 g/mol. A CSF glucose of 60 mg/dL is 3.33 mmol/L. To go the other way, divide.

Q: What is a normal CSF glucose?

ARUP publishes 50–80 mg/dL (2.78–4.44 mmol/L) for cerebrospinal fluid; Labcorp publishes 49–73 mg/dL for over-17s and lower bands for children, down to 41–59 mg/dL under 5 years. The two upper limits differ by about 10%, so check which laboratory issued your result.

Q: Why does the CSF glucose need a blood glucose with it?

Because CSF glucose is about 60% of plasma glucose and follows it with a one- to three-hour lag, so the absolute value is only interpretable against the plasma value at the time of the tap. A CSF glucose of 70 mg/dL is normal if the plasma is 110 and markedly abnormal if the plasma is 300. The blood sample must be drawn simultaneously, and before any glucose infusion is started.

Q: What CSF glucose suggests bacterial meningitis?

Labcorp's interpretive note on test 002048 gives 30 mg/dL as typical of bacterial meningitis and 20 mg/dL or less as highly correlated with it. Both are absolute figures and both assume a normal plasma glucose; the ratio is the more robust reading, and a normal glucose does not exclude a partially treated infection or a viral cause.

Q: Is CSF glucose ever high?

Only as a reflection of hyperglycaemia at the time of the tap. There is no central nervous system process that raises CSF glucose independently, so a high value is a comment on the blood rather than on the brain.

Convert CSF glucose between mg/dL, mg/L, g/L and mmol/L — and see why the number means little without the plasma glucose drawn beside it.

Last updated 8 Oct 2026

CSF Glucose converter

Mass ⇄ molar

CSF glucose

A CSF glucose on its own is close to uninterpretable. What the test actually is, is the ratio to a plasma glucose drawn at the same time.

Reference sourceARUP and Labcorp do not agree at the top of the adult interval; pick the one that issued your result.

3.33mmol/LExample

CSF glucose 60 mg/dL, with a plasma glucose of 100 mg/dL drawn at the same time

The conversion, and the number that actually matters

mmol/L = mg/dL × 0.0555074
mg/dL = mmol/L ÷ 0.0555074
mg/L = mg/dL × 10
CSF:plasma glucose ratio = CSF glucose ÷ simultaneous plasma glucose — unitless, so it is the same number in either system

180.156: the molar mass of glucose, C6H12O6. The same molecule as plasma glucose, so the same factor as the serum converter — only the specimen and the reference interval differ
0.0555: the mg/dL to mmol/L factor, identical to the one on every other glucose page on this site
the ratio: dimensionless, which is the one practical advantage it has over the absolute value: a reader in mmol/L and a reader in mg/dL compute the same number and compare it against the same threshold

Worked example

CSF glucose 60 mg/dL, with a plasma glucose of 100 mg/dL drawn at the same time
60 × 0.0555074 = 3.33 mmol/L
60 mg/dL = 600 mg/L = 0.600 g/L
Ratio = 60 ÷ 100 = 0.60, which is normal — and would still be 0.60 if both numbers were in mmol/L
Inside ARUP's 50–80 mg/dL and inside Labcorp's 49–73 mg/dL for an adult, so the absolute value is unremarkable on either interval

CSF glucose: what the two laboratories actually publish

ARUP publishes one adult interval and Labcorp publishes five age bands, and they disagree at the top: a CSF glucose of 76 mg/dL is normal at ARUP and above the interval at Labcorp, a 10% difference in the upper limit for the same molecule. Labcorp alone separates boys and girls between 6 and 12 years. Reference intervals are method- and laboratory-dependent, and the interval printed on your own report takes precedence over any figure here.
Source	Interval (mg/dL)	Same interval (mmol/L)
ARUP 3016615, cerebrospinal fluid	50 – 80	2.78 – 4.44
Labcorp 002048, over 17 years	49 – 73	2.72 – 4.05
Labcorp 002048, 13-17 years	45 – 66	2.50 – 3.66
Labcorp 002048, 6-12 years, male	46 – 62	2.55 – 3.44
Labcorp 002048, 6-12 years, female	43 – 60	2.39 – 3.33
Labcorp 002048, 0-5 years	41 – 59	2.28 – 3.27

The thresholds people actually use, and they are ratios

Both absolute thresholds come from Labcorp’s own interpretive note on test 002048. Neither survives a plasma glucose that is itself abnormal, which is why the ratio rather than the value is what the guidelines are written around — and why CSF lactate, which needs no paired blood sample at all, outperforms the ratio in practice.
Finding	Threshold	Source
CSF glucose strongly suggesting bacterial meningitis	≤ 20 mg/dL (1.11 mmol/L)	Labcorp 002048
CSF glucose typical of bacterial meningitis	< 30 mg/dL (1.67 mmol/L)	Labcorp 002048
CSF:plasma glucose ratio, normal	about 0.6	see the ratio interpreter
Absolute value with a hyperglycaemic plasma	unreliable — use the ratio	—

Why the CSF glucose is not really a CSF test

Glucose reaches the cerebrospinal fluid by facilitated transport across the blood–brain barrier, and the CSF concentration tracks the plasma concentration with a lag of one to three hours at roughly 60% of it. Both halves of that sentence matter. The proportionality is why the test works at all: organisms and leucocytes consume glucose, and impaired transport across an inflamed barrier compounds it, so bacterial, tuberculous and fungal meningitis drive the CSF glucose down. The lag is why the absolute number misleads so often.

Two failure modes follow directly. A diabetic patient with a plasma glucose of 300 mg/dL and florid bacterial meningitis can present with a CSF glucose of 70 mg/dL — inside every interval on this page, and a ratio of 0.23. A hypoglycaemic patient with a perfectly normal CSF can present with a CSF glucose of 30 mg/dL. The ratio corrects both, and because it is dimensionless it is the one number in CSF chemistry that does not need converting at all.

The lag is the reason the blood sample must be drawn at the same time as the tap, not afterwards and not after an infusion has started — and that sample is the one most often missing from an out-of-hours workup. Where it is missing, CSF lactate does the same job on a single tube, and in the published cohorts it does it better.

A third trap is specific to the unit on this page. CSF glucose is reported in mg/dL across the United States and in mmol/L across most of Europe, and the two differ by a factor of about 18. A threshold read in the wrong system is not slightly wrong; 2.2 mmol/L and 2.2 mg/dL are a normal result and an unsurvivable one. The molecule is the same as on the serum glucose converter, so the factor is the same — only the interval and the meaning change.

Frequently asked questions

How do I convert CSF glucose from mg/dL to mmol/L?

Multiply by 0.0555074, the same factor as for plasma glucose, because it is the same molecule with a molar mass of 180.156 g/mol. A CSF glucose of 60 mg/dL is 3.33 mmol/L. To go the other way, divide.

What is a normal CSF glucose?

ARUP publishes 50–80 mg/dL (2.78–4.44 mmol/L) for cerebrospinal fluid; Labcorp publishes 49–73 mg/dL for over-17s and lower bands for children, down to 41–59 mg/dL under 5 years. The two upper limits differ by about 10%, so check which laboratory issued your result.

Why does the CSF glucose need a blood glucose with it?

Because CSF glucose is about 60% of plasma glucose and follows it with a one- to three-hour lag, so the absolute value is only interpretable against the plasma value at the time of the tap. A CSF glucose of 70 mg/dL is normal if the plasma is 110 and markedly abnormal if the plasma is 300. The blood sample must be drawn simultaneously, and before any glucose infusion is started.

What CSF glucose suggests bacterial meningitis?

Labcorp’s interpretive note on test 002048 gives 30 mg/dL as typical of bacterial meningitis and 20 mg/dL or less as highly correlated with it. Both are absolute figures and both assume a normal plasma glucose; the ratio is the more robust reading, and a normal glucose does not exclude a partially treated infection or a viral cause.

Is CSF glucose ever high?

Only as a reflection of hyperglycaemia at the time of the tap. There is no central nervous system process that raises CSF glucose independently, so a high value is a comment on the blood rather than on the brain.

Related calculators

References

ARUP Laboratories. CSF, glucose — reference interval. Laboratory Test Directory 3016615; 2026.
Labcorp. Glucose, cerebrospinal fluid — test 002048, age-stratified reference intervals and interpretive thresholds for bacterial meningitis. Test menu; 2026.
Commission on Isotopic Abundances and Atomic Weights, IUPAC. Standard Atomic Weights 2024. Accessed October 2026.

Not medical advice. For healthcare professionals and education. Reference intervals vary by laboratory and assay — always use your own laboratory's. Never base a dose or a treatment decision on this page alone. Full disclaimer at calcengines.com/disclaimer/