CSF Albumin Unit Converter

Convert CSF albumin between mg/L, mg/dL, g/L, µg/mL and µmol/L — and get the unit right before it reaches the albumin quotient, where a factor of ten hides in plain sight.

Last updated 8 Oct 2026

CSF Albumin converter

Mass ⇄ molar

CSF albumin

Mayo and ARUP both print this in mg/dL. The Q-albumin formula takes it in mg/L. That thousand-fold mismatch is the single commonest arithmetic error in CSF reporting.

Reference sourceMayo and ARUP disagree by 30% at the upper limit; pick the one that issued your result.

24.00mg/dLExample

CSF albumin reported as 240 mg/L, with a serum albumin of 42 g/L

The conversion, and the unit trap in the quotient

mg/dL = mg/L ÷ 10
g/L = mg/L ÷ 1000
µg/mL = mg/L × 1 — the same quantity written two ways
µmol/L = mg/L × 0.0150376
Q-albumin = CSF albumin (mg/L) ÷ serum albumin (g/L)

66,500: the molar mass of human serum albumin, UniProt P02768. The same protein as in serum, so the same factor as the serum converter — it is the magnitude of the number, not the molecule, that changes between the two specimens
mg/L against g/L: the quotient deliberately mixes them, which is why its value is around 5 rather than around 0.005. Feed it a CSF albumin in mg/dL and the quotient comes out ten times too low, and a leaking barrier reads as intact
µg/mL: numerically identical to mg/L. A report in µg/mL needs no arithmetic at all before it goes into the quotient

Worked example

CSF albumin reported as 240 mg/L, with a serum albumin of 42 g/L
240 ÷ 10 = 24.00 mg/dL, which is the unit Mayo and ARUP print
240 mg/L = 240 µg/mL = 0.2400 g/L = 3.609 µmol/L
Q-albumin = 240 ÷ 42 = 5.7 — an intact barrier for an adult
Had the 24 mg/dL figure gone into the quotient instead, it would have read 0.57, and the same patient would have looked impossibly watertight

CSF albumin: the two laboratories, in both units

ARUP’s upper limit for CSF albumin is 30% above Mayo’s, so a CSF albumin of 30 mg/dL is raised at Mayo and normal at ARUP. Note what the two serum rows do to the quotient: both laboratories print serum albumin in mg/dL as well, and the quotient wants it in g/L. Reference intervals are method- and laboratory-dependent, and the interval printed on your own report takes precedence over any figure here.
Source	As published (mg/dL)	Same interval (mg/L)	And in µmol/L
Mayo SFIG, CSF albumin	0.0 – 27.0	0 – 270	0 – 4.06
ARUP 0050676, CSF albumin	0 – 35	0 – 350	0 – 5.26
Serum albumin, for the quotient Mayo SFIG	3,500 – 5,000	35,000 – 50,000	—
Serum albumin ARUP 0050676	3,500 – 5,200	35,000 – 52,000	—

Where this number goes next, and in which unit

Four calculators on this site take a CSF albumin and three different unit conventions between them, which is exactly why this page exists. The IgG index is immune to the error because the same unit cancels top and bottom; the quotient is not, because it deliberately does not cancel. Mayo’s own upper limit for the quotient is 14, against the 7 commonly quoted from Reiber’s work for a young adult — read it against an age-specific limit, not a single figure.
Calculation	Wants CSF albumin in	Wants serum albumin in
Q-albumin	mg/L	g/L
IgG index	mg/dL	mg/dL
IgG synthesis rate	mg/dL	mg/dL
Reiber quotient diagram	mg/L	g/L

What CSF albumin measures, and why its unit is a hazard

Albumin is not synthesised anywhere in the central nervous system. Every molecule of it in cerebrospinal fluid arrived from blood, which makes it the cleanest available probe of the blood–CSF barrier: the CSF concentration divided by the serum concentration is a pure measure of how much protein is getting across, with no intrathecal production term to subtract. That is the whole basis of the albumin quotient, and it is why albumin rather than total protein is the parameter Reiber’s tradition is built on.

What makes the result difficult is arithmetic rather than physiology. CSF albumin is about a two-hundredth of serum albumin, so the two are conventionally reported in units a thousand apart — CSF in mg/L, serum in g/L — and the quotient is defined to exploit that mismatch, which is why a normal Q-albumin is a number around 5 rather than around 0.005. Both Mayo and ARUP, however, print CSF albumin in mg/dL. Put Mayo’s 24 mg/dL straight into the quotient instead of its 240 mg/L equivalent and the answer is ten times too low: a patient with Guillain-Barré syndrome and a frankly leaking barrier is reported as having an intact one.

The second trap is that there is no single upper limit. Barrier permeability rises with age, and the published limit rises with it — from about 5 under 15 years to around 9 or 10 over 60. Mayo prints a flat upper limit of 14 for its own quotient, which is higher than any of those. A mildly raised quotient in a 70-year-old may be entirely normal for their age and frankly abnormal for a student, and the only way to tell is to use an age-stratified limit. The quotient calculator carries the age bands; this page only makes sure the number entering it is in the unit it expects.

The molar column is included because albumin, unlike CSF total protein or CSF IgG, genuinely has one. It is a single defined polypeptide of 66,500 Da, so dividing by that mass is a real conversion rather than a convention. Almost no laboratory reports CSF albumin in µmol/L, but if one does, the figure is directly comparable with the µmol/L column on the serum albumin converter.

Frequently asked questions

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

Multiply by 10. Mayo’s upper limit of 27.0 mg/dL is 270 mg/L and ARUP’s 35 mg/dL is 350 mg/L. This is the conversion to get right before the number reaches the albumin quotient, which takes CSF albumin in mg/L and serum albumin in g/L.

What is a normal CSF albumin?

Mayo publishes 0.0–27.0 mg/dL (0–270 mg/L) and ARUP publishes 0–35 mg/dL (0–350 mg/L). The 30% disagreement at the upper limit is real and not a rounding artefact, so a CSF albumin of 30 mg/dL is raised at one laboratory and normal at the other.

Why is CSF albumin in mg/L when serum albumin is in g/L?

Because CSF albumin is roughly two hundred times lower, so mg/L keeps it as a readable whole number while g/L does the same for serum. The albumin quotient then deliberately mixes the two, which is the only reason a normal Q-albumin is about 5 rather than about 0.005.

Can CSF albumin be converted to µmol/L?

Yes. Albumin is a single defined protein of 66,500 Da (UniProt P02768), so 1 mg/L is 0.0150376 µmol/L. This is the difference between albumin and CSF total protein or CSF IgG, both of which are mixtures with no single molar mass and are therefore reported in mass units only.

Is a raised CSF albumin the same as a raised CSF protein?

Usually but not always, and they answer different questions. Albumin is purely blood-derived, so the quotient isolates barrier function. CSF total protein includes locally produced immunoglobulin, so a raised protein with a normal albumin quotient points to intrathecal synthesis rather than a leaking barrier — which is what the IgG index is built to detect.

Related calculators

References

Mayo Clinic Laboratories. Cerebrospinal fluid IgG index profile, serum and spinal fluid (SFIG) — reference values for CSF IgG, CSF albumin, the IgG index, the IgG synthesis rate and the albumin quotient. Test catalogue, test ID 610783; 2026.
ARUP Laboratories. Immunoglobulin G, CSF index — reference intervals for CSF IgG, CSF albumin, serum IgG and serum albumin, the IgG index and the IgG synthesis rate. Laboratory Test Directory 0050676; 2026.
UniProt. P02768, human serum albumin — mature protein 585 residues, about 66.5 kDa. 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/