What GFR actually measures
Your kidneys contain roughly a million filtering units called nephrons. Blood arrives at each one under pressure, and a filter — the glomerulus — lets water and small solutes through while holding back cells and large proteins. Glomerular filtration rate is the total volume of that filtrate produced per minute across both kidneys, normalised to a standard body surface area of 1.73 m². It is the single best summary of how much working kidney you have.
True GFR is measured by injecting a marker such as iohexol or inulin and tracking its clearance — accurate, expensive, and rarely done. So instead we estimate it from an endogenous marker already in your blood. Creatinine is the standard one: muscle produces it at a fairly steady rate, and the kidneys clear it. If clearance falls, creatinine rises. The equations exist to convert that rise back into a filtration rate, correcting for the fact that creatinine production varies with age, sex and body composition.
The CKD-EPI 2021 creatinine equation
The two-part creatinine term is doing something clever: the min() branch applies a shallow exponent while creatinine is below the sex-specific reference κ, and the max() branch applies a much steeper −1.200 once it goes above. That is why the curve in the calculator is steep at the low end and flattens at the high end — a rise from 0.9 to 1.1 mg/dL costs far more eGFR than a rise from 2.4 to 2.6.
The 2021 race-free change — what actually happened
Until 2021, both the MDRD and CKD-EPI equations carried a race coefficient. In CKD-EPI 2009 it multiplied eGFR by 1.159 — about 16% higher — if the patient was recorded as Black. That was not ideology; it came from the data. In the development cohorts, Black participants had, on average, higher serum creatinine at the same measured GFR, most plausibly because of average differences in muscle mass and creatinine generation between the groups as sampled.
The problem is what the coefficient was standing in for. Race is a social category, not a measurement of anybody's muscle mass, and applying a population-average multiplier to an individual produced a systematic bias: the same blood result reported better kidney function for a Black patient than for anyone else. Downstream, that meant later CKD diagnosis, later nephrology referral, later drug-dose adjustment, and — because transplant waitlist eligibility keys off eGFR — later listing for a kidney.
A joint National Kidney Foundation and American Society of Nephrology task force worked through 2020–21 and recommended dropping race entirely. Inker and colleagues refitted the equations without it (published in the New England Journal of Medicine, November 2021), and the result is what this calculator uses. The refit is not merely the old equation with the coefficient deleted: every parameter was re-estimated. In practice eGFR values came down modestly for Black adults and rose modestly for everyone else, and accuracy in the population as a whole was essentially preserved. The task force also recommended much wider use of cystatin C, which is the real fix for the underlying problem.
The KDIGO categories
| Category | eGFR | Meaning |
|---|---|---|
| G1 | ≥ 90 | Normal or high |
| G2 | 60–89 | Mildly decreased |
| G3a | 45–59 | Mildly to moderately decreased |
| G3b | 30–44 | Moderately to severely decreased |
| G4 | 15–29 | Severely decreased |
| G5 | < 15 | Kidney failure — dialysis or transplant territory |
Two things about this table are widely misread. First, G1 and G2 are not chronic kidney disease by themselves. CKD requires either eGFR under 60 or a marker of kidney damage — most often albuminuria — persisting for three months or longer. A single eGFR of 75 in an otherwise healthy person is not a diagnosis. Second, full KDIGO staging is two-dimensional: the G category above, paired with an albuminuria category A1 to A3. Risk depends on both, and a person with G2 plus heavy albuminuria can be at higher risk than someone with G3a and none.
What moves creatinine besides your kidneys
Every source of variation in creatinine production shows up as false variation in estimated filtration:
- Muscle mass. More muscle, more creatinine, lower apparent eGFR. Strength athletes and bodybuilders are routinely flagged. In the other direction, low muscle mass from age, malnutrition, cirrhosis, amputation or paralysis makes eGFR read falsely high — which is the more dangerous error, because it hides real disease.
- Creatine supplements. Supplemental creatine raises serum creatinine without touching kidney function.
- A large meat meal. Cooked meat contains creatinine directly; a big steak can lift a result measurably for hours.
- Hydration. Dehydration lowers renal perfusion and raises creatinine; a well-hydrated repeat draw often looks better.
- Drugs. Trimethoprim and cimetidine block creatinine secretion in the tubule, raising the number with no change in filtration. Many other drugs affect the kidney itself.
- Pregnancy and acute illness. Both change kidney physiology enough that steady-state equations do not apply.
Cystatin C: the alternative marker
Cystatin C is a small protein produced by essentially all nucleated cells and cleared almost entirely by glomerular filtration. Because production does not depend on muscle, it sidesteps the biggest source of error in creatinine-based estimates. KDIGO and the NKF–ASN task force both recommend using it to confirm eGFR when the result would change management — before starting a nephrotoxic drug, when assessing transplant candidacy, or in anyone whose body composition makes creatinine unreliable. The combined creatinine + cystatin C equation (CKD-EPI 2021 cr-cys) is more accurate than either marker alone and is what to ask about if your creatinine-based number and your clinical picture disagree.
Reading your result sensibly
Look at the trend, not the point. A stable eGFR of 62 over four years is a very different situation from a fall of 95 to 62 in eighteen months, even though the current number is identical. Ask for a urine albumin-to-creatinine ratio alongside it, because albuminuria carries independent risk and is cheap to measure. And treat this page for what it is: arithmetic on one blood value. It cannot see your urine, your blood pressure, your medication list or your history — all of which a clinician needs before a number becomes a diagnosis.
While you are here: blood pressure and metabolic health are the two biggest modifiable inputs to long-term kidney function, and body composition is what makes creatinine-based estimates go wrong. Our BMI calculator and body fat calculator give the context; the BAC calculator covers the other organ that quietly does a lot of filtering.