GFR Calculator – eGFR by CKD-EPI 2021, With Your CKD Stage

Where your result sits — the KDIGO G-stages

Clinicians sort eGFR into six bands called G-stages. Find your number on the gauge, then read your row. One caution that matters more than any single number: a stage is a category for a result, not a diagnosis — chronic kidney disease requires the finding to persist for at least three months, or other markers of kidney damage to be present.

Keep exploring

How your eGFR is worked out, in three steps

The equation, and every assumption in the open

Going deeper

The same creatinine, six different answers

Creatinine alone does not determine eGFR. The CKD-EPI 2021 equation reads the same blood result differently by sex and by age, because the muscle mass that produces creatinine and the slow decline of filtration with age are both built into the maths. The table below holds creatinine at 1.2 mg/dL and varies only sex and age. The 60-year-old rows were produced by the calculator itself, and the 30-year-old rows follow the same published equation and match the calculator’s output.

Age alone moves the 1.2 result by 14 points for a man (83.4 at 30 against 69.2 at 60) and by 11 points for a woman (62.5 against 51.8), because the age factor of 0.9938 per year compounds: 0.9938 to the power 30 is about 0.830, while to the power 60 it is 0.689. Sex moves it more: at 60, the same creatinine gives 69.2 for a man and 51.8 for a woman — a gap of 17 points, enough to cross from G2 into G3a — because a woman’s κ is 0.7 rather than 0.9, so her Scr/κ ratio is 1.71 rather than 1.33, and the steeper part of the curve bites harder. None of this means the woman’s kidneys filter less well in any absolute sense. It means the same creatinine implies different filtration once the muscle mass behind it is accounted for, which is exactly what the sex-specific κ is for. Comparing raw creatinine values between two people without age and sex is therefore meaningless, and comparing eGFR values across the two equations on this page without noting which equation produced them is nearly as bad.

Why the equation bends: the kink at κ

The CKD-EPI 2021 equation is two curves joined at κ. Below κ, the ratio Scr/κ is raised to a shallow power — α, which is −0.302 for men and −0.241 for women. Above κ, it is raised to a steep −1.200 for everyone. That kink is what lets one equation track both healthy kidneys and failing ones, and it has a practical consequence: eGFR moves slowly when creatinine is low and fast when creatinine is high. Take a 60-year-old man, whose κ is 0.9, and walk his creatinine up. Every figure below was produced by the calculator itself.

Below κ, the curve is forgiving: dropping creatinine from 0.9 to 0.6 lifts eGFR from 97.8 to 110.5, a gain of 12.7 points for a 0.3 change. Above κ, the same 0.3 change costs far more: rising from 0.9 to 1.2 costs 28.6 points, and the next 0.1, from 1.2 to 1.3, costs another 6.3 points (69.2 down to 62.9). That asymmetry explains the most common confusion on this page: a creatinine that “barely moved” can move eGFR a lot, provided the move happened above κ, where the steep power rules. It also explains why the equation needs the kink at all. A single straight power would either overstate filtration in healthy kidneys or understate it in failing ones; the shallow-then-steep shape follows measured GFR on both sides of the reference point instead.

2021 against 2009, and the race coefficient, in numbers

The comparison panel on this page exists because laboratories changed equations, and the change moved reported numbers without any change in the patient. For the default 60-year-old man at a creatinine of 1.2 mg/dL, the 2021 equation gives 69.2 while the retired 2009 equation without its race coefficient gives 65.3 — the 2021 refit reads 3.9 points higher. For a 60-year-old woman at the same creatinine, the figures are 51.8 against 49.1, a gap of 2.7 points. Both gaps come from the refit’s rebuilt constants: 142 instead of 141, an age factor of 0.9938 per year instead of 0.993, and new α values.

The bigger shift belongs to patients whose old reports included the 2009 race coefficient of 1.159. On the same default man, the 2009 figure with the coefficient is 75.7 — the calculator shows this when the comparison selector is set to the old coefficient — against 69.2 from the 2021 equation, a fall of 6.5 points with no change in blood results at all. The direction matters: the 2021 refit reads higher than the 2009 equation without the coefficient, but lower than the 2009 equation with it. So if your new report is a few points above your old one, the equation change is the first suspect; if it is several points below an old report that carried the coefficient, the removal of the coefficient is the likely explanation. In both cases the kidneys did not necessarily change — the ruler did. The 2021 equation was recommended by the National Kidney Foundation and American Society of Nephrology task force precisely so that one rule serves everyone, and the combined creatinine and cystatin C equation published alongside it recovers accuracy without race for cases where a single-marker estimate is not trusted.

What moves creatinine besides filtration

An eGFR is an estimate worked out from creatinine, not a measurement of filtration, and creatinine moves for reasons that have nothing to do with the kidneys. The disclaimer on this page names the main ones: muscle mass, a recent cooked-meat meal, creatine supplements, and some medicines. A heavy training week, dehydration or a recent illness can pull a single result down temporarily, and the stage section on this page makes the same point about a one-off dip: it deserves a repeat test before it deserves worry. That is not a flaw in the equation. It is the nature of a marker that is produced by muscle and cleared by the kidneys — anything that changes production or volume status changes the reading while filtration itself stands still.

The practical handling follows from that. A single result is a reading of a lab value, not a verdict on your kidneys, and chronic kidney disease as a diagnosis requires the finding to persist for at least three months, or damage markers such as albumin in the urine alongside it. If a result surprises you, the first questions are the boring ones: were you dehydrated, ill, or training hard; had you eaten a large cooked-meat meal or started creatine; has any medicine changed; and does the new report use the 2021 equation where the old one used 2009? If the result repeats, or if urine markers are present, that is the point to discuss it with your doctor, who can check the trend, the urine test that completes the picture, and whether a cystatin C measurement would help. This page gives you the arithmetic and the context for that conversation. It does not diagnose, and no single number on it should be read as one.

MDRD: the equation this page replaced

Before CKD-EPI, most laboratories reported eGFR from the MDRD equation, built from the Modification of Diet in Renal Disease study and refitted to standardised creatinine assays in 2006. It takes the same three inputs as the equation on this page, creatinine, age, and sex, but it bends them differently, and on the same blood result it usually reads lower. The table below compares the two equations with no race coefficient applied to either, so the difference is the maths alone.

The MDRD equation, in its standardised form, is: eGFR = 175 × Scr^−1.154 × Age^−0.203 × 0.742 if female. There is no kink at a reference creatinine, as in CKD-EPI, and no separate sex constant dividing creatinine first: one power on creatinine and one power on age carry the whole estimate. The retired MDRD race coefficient, 1.212, worked the same way as the retired 2009 CKD-EPI coefficient, and it is not applied anywhere in the table.

Two patterns in that table explain why MDRD was retired. First, the gap is largest where filtration is highest: 12.3 points for a 30-year-old man and 11.7 points at a low creatinine of 0.9, but only 3.2 points once creatinine reaches 2.0. MDRD was fitted mainly on people who already had kidney disease, so it tracks low filtration well and understates healthy filtration, which is why many laboratories reported MDRD results above 60 only as above 60 rather than as a number. Second, the gap shrinks with age in these examples, because the two equations age differently. CKD-EPI was fitted across the full range, from healthy volunteers to people with advanced disease, which is why it replaced MDRD for routine reporting, and the 2021 refit then removed the race term that both older equations carried.

If an old report of yours says MDRD, do not compare its number directly with the number on this page. Expect the CKD-EPI 2021 figure to read higher, especially if the old result was capped at above 60 or came from a younger age. The comparison panel on this page shows the 2009 CKD-EPI equation rather than MDRD, because 2009 was the equation most laboratories actually used in the years immediately before the 2021 switch. MDRD belongs to the report before that, and this section exists so that an even older number is not mistaken for a change in your kidneys.

Your eGFR by age: the same creatinine, seven decades

Age is built into the equation, so the same creatinine means a different eGFR at 20 and at 80. The table holds creatinine fixed at 1.2 mg/dL and walks through the decades, using the CKD-EPI 2021 equation exactly as this calculator does. It is arithmetic, not a table of expected values: real creatinine varies from person to person, and a healthy older adult often has a lower creatinine than 1.2.

At a fixed creatinine, the equation trims roughly 4 to 5 points per decade, because the age factor of 0.9938 per year compounds: a man at 1.2 mg/dL reads 83.4 at 30, 69.2 at 60, and 61.1 at 80. The sex gap is wider than the decade gap at every age in this table, 22.3 points at 20 and 15.3 points at 80, because the sex-specific constant divides creatinine before anything else happens. That is also why the stage column looks stricter for women at the same creatinine: the stage follows the estimate, and the estimate follows the muscle mass the creatinine came from. None of this sets a normal value for any individual. A result below 60 draws clinical attention, and a result in the 60s or 70s can be entirely ordinary at older ages, but only the trend across repeat tests, together with urine markers, turns a number into a clinical picture. This page is general information, not a diagnosis, and a result that worries you belongs in a conversation with your doctor.

Cystatin C: the second marker, for when creatinine does not fit the person

Creatinine is a muscle story, as the guide above explains, so when muscle mass is very high, very low, or changing fast, the creatinine estimate can mislead in either direction. Cystatin C is a second blood marker that is made by nearly all cells and cleared by the kidneys, and it depends far less on muscle. The CKD-EPI group published equations based on cystatin C alone in 2012, and combined creatinine and cystatin C equations in 2012 and again in the 2021 refit, which recover accuracy without using race. This calculator is creatinine-only and does not estimate from cystatin C: if the creatinine figure looks unrepresentative for the person in front of you, a cystatin C based estimate ordered through your doctor is the usual next step, not a different creatinine equation.

Reading the blood test behind the number: creatinine, BUN, and the rest of the panel

An eGFR never arrives alone. On a laboratory report it sits beside the creatinine result it was calculated from, and usually beside several other kidney-related lines, and each line answers a different question. Creatinine is the input: a waste product from muscle, measured in blood serum, reported in mg/dL in the United States and in micromol/L in most other countries, where the two units convert exactly, micromol/L divided by 88.4 gives mg/dL. The equation on this page reads that one number, with age and sex, and returns the estimate. Blood urea nitrogen, BUN, is a different waste product, urea, and it is not an input to the CKD-EPI equation: a high BUN with a normal creatinine points a clinician toward causes outside filtration itself, such as dehydration, a high-protein meal, or bleeding in the gut, while a high creatinine with a normal BUN points more directly at filtration. The two numbers are read together precisely because they move for different reasons, and neither one, alone, diagnoses anything.

Two habits make a blood-test result easier to use. First, compare like with like: the same laboratory, the same units, and the same equation. Laboratories changed equations over the past decade, from MDRD to CKD-EPI 2009 and then to CKD-EPI 2021, and the change alone moves the reported number by several points with no change in the patient, which is why an old report and a new report can disagree while the kidneys have done nothing at all. Second, treat one result as one measurement, not as a verdict. Creatinine moves with muscle mass, with a recent cooked-meat meal, with creatine supplements, with dehydration, and with some medicines, so a single low eGFR is repeated and confirmed before it is called chronic: under the KDIGO 2024 guideline, chronic kidney disease means an abnormality present for at least three months, shown by past measurements, by repeat testing, or by other evidence of kidney damage. A result below 60 that persists, or a result at any level with albumin in the urine, is the finding to take to a clinician promptly, with the full panel in hand, rather than a single number on its own.

When creatinine alone is not enough: the cystatin C confirmatory estimate

Creatinine has a weakness built into it: it is produced by muscle, so anything unusual about muscle mass makes the estimate less trustworthy. A very muscular person produces more creatinine at the same filtration, and the equation reads it as worse kidney function than is really there; a person with very low muscle mass, an amputee, or someone with a long illness produces less, and the equation reads it as better function than is really there. Cystatin C is a second blood marker that avoids most of that problem: it is produced by nearly all nucleated cells at a fairly steady rate, so it tracks filtration with far less dependence on muscle. The KDIGO 2024 guideline therefore recommends that, where cystatin C is available, the GFR category be estimated from the combination of creatinine and cystatin C, the eGFRcr-cys estimate, rather than from creatinine alone, and the combined estimate is the usual confirmatory step when a creatinine-based result looks unrepresentative for the person in front of the clinician.

This calculator is creatinine-only, and that limit is stated on the page for a reason: a cystatin C estimate needs a cystatin C result, which is a separate laboratory test, ordered through a doctor, and it is not interchangeable with the number here. A combined estimate will usually sit close to the creatinine estimate when muscle mass is average, and further from it when muscle mass is far from average, in either direction, and the gap itself is the information: a large gap says the creatinine figure was being pulled by something other than filtration. Do not convert one into the other by hand, and do not average the two numbers yourself: the combined equation weights the two markers in its own way, and the result belongs with the clinician who ordered the test, alongside the urine albumin result and the repeat measurement that establishes whether the finding has persisted for three months.

Frequently asked questions

Related guides

The equations compared, and your number read by age.

Related calculators