Last updated: October 2026
By CalcOrigin Editorial Team
What GFR Measures
Glomerular filtration rate is the volume of plasma the kidneys filter in a given time. The glomeruli are the capillary tufts where filtration begins, and together they process roughly 180 litres of plasma a day in a healthy adult — about 125 millilitres a minute. Almost all of that volume is reabsorbed downstream, which is why urine output is measured in litres rather than in the same units.
GFR is usually indexed to body surface area so that people of different sizes can be compared. That is what the units mean: millilitres per minute per 1.73 square metres, the surface area of an average adult. A result of 90 is therefore not a claim about the absolute volume filtered by that individual, but about their filtration relative to a standard body size.
Values fall with age even in people without kidney disease. Filtration peaks in the third decade and then declines slowly, on the order of 1 mL/min/1.73 m² per year after about 40. A result of 75 in a healthy 75-year-old may be entirely unremarkable, while the same number in a 30-year-old warrants investigation.
Because the decline is slow and symptomless, a single value says less than a trend. A stable 55 that has been 55 for three years is a different finding from a value that was 90 eighteen months ago. This is why clinicians ask for previous results before interpreting a new one, and why the calculator result should be read alongside history rather than in isolation.
GFR is also only one half of kidney staging. Albumin in the urine reflects damage to the filtration barrier and often appears before filtration rate falls. Someone with a normal GFR and persistent albuminuria still has kidney disease, and the risk of progression depends on both numbers together.
This page estimates GFR from a blood creatinine result and a few personal details. It cannot measure filtration directly, and it cannot diagnose anything. Its purpose is to put a number on a laboratory report in context and to help you ask better questions of the clinician who ordered the test.
Typical Values by Age
Reference values are averages, not cutoffs. The table shows roughly where estimated GFR sits in adults without known kidney disease, using the age bands commonly reported in population data. Individual values scatter widely around these figures for reasons that have nothing to do with disease.
| Age |
Approximate mean eGFR |
| 20–29 | 116 mL/min/1.73 m² |
| 30–39 | 107 |
| 40–49 | 99 |
| 50–59 | 93 |
| 60–69 | 85 |
| 70 and over | 75 |
The staging thresholds used in clinical practice are fixed rather than age-adjusted, which creates an awkward overlap: a healthy 80-year-old can fall below 60 and be labelled stage 3a without having a disease process at all. Guidelines address this by requiring other evidence — persistent albuminuria, structural abnormality, or a sustained decline — before a diagnosis of chronic kidney disease is made on the basis of filtration rate alone.
Muscle mass is the largest single source of variation in creatinine-based estimates. A muscular young man may have a creatinine of 1.2 mg/dL and a normal GFR, while a frail older woman with the same creatinine may have a considerably reduced one. The equation corrects for age and sex but only crudely, which is why cystatin C is preferred in people at the extremes of body composition.
Pregnancy raises GFR substantially, by up to half above pre-pregnancy values, and creatinine falls accordingly. Standard equations were not derived in pregnancy and should not be used to interpret a pregnant patient's result without specialist input.
Finally, laboratory assays differ. Creatinine measured by an IDMS-traceable enzymatic method is not identical to older picrate-based methods, and the estimating equations assume the traceable standard. A result that shifts by a few percent after a laboratory changes its assay may reflect the assay rather than the kidney.
Stages and Albuminuria Categories
Chronic kidney disease is staged by filtration rate, with the third stage split in two because the difference in risk between 50 and 35 is substantial. The letter G refers to the GFR category; a separate letter A grades albuminuria. Both are needed to describe severity.
| Category |
eGFR, mL/min/1.73 m² |
Description |
| G1 | 90 or above | Normal or high, but CKD requires evidence of damage |
| G2 | 60–89 | Mildly reduced, again requiring damage markers for CKD |
| G3a | 45–59 | Mild to moderately reduced |
| G3b | 30–44 | Moderately to severely reduced |
| G4 | 15–29 | Severely reduced |
| G5 | Below 15 | Kidney failure; dialysis or transplant may be needed |
Albuminuria is graded from the urine albumin-to-creatinine ratio, which corrects a spot urine sample for dilution and avoids a 24-hour collection.
| Category |
Albumin-to-creatinine ratio |
Term |
| A1 | Below 30 mg/g | Normal to mildly increased |
| A2 | 30–300 mg/g | Moderately increased (formerly microalbuminuria) |
| A3 | Above 300 mg/g | Severely increased (formerly macroalbuminuria) |
Risk rises with both axes, and the combination matters more than either number alone. A G2A3 result — near-normal filtration with heavy albuminuria — carries a higher risk of progression than a G3aA1 result, because albuminuria signals active damage to the glomerular barrier rather than the slow loss of filtration that accompanies age.
The definition of CKD requires abnormality present for at least three months. One low reading does not qualify, because acute illness, dehydration, and a heavy protein meal can all depress filtration temporarily. Confirmation on a repeat sample after an interval is part of the diagnosis, not an optional extra.
Management intensity follows the grid. G1 and G2 with A1 usually need only annual review and control of blood pressure and glucose. G3 with A2 or A3 warrants more frequent monitoring, blood pressure treatment, and consideration of medications that slow progression. G4 and G5 need specialist care and preparation for kidney replacement therapy.
Measured Versus Estimated
The reference method infuses a substance that is filtered freely, is neither secreted nor reabsorbed by the tubules, and is not metabolized by the kidney. Inulin fits those criteria, and inulin clearance remains the historical standard. Iohexol, iothalamate, and radiolabelled EDTA or DTPA are used in modern research and in selected clinical situations.
These measurements are accurate but awkward. They require an intravenous infusion or injection, timed blood and urine collections, and laboratory support. The result is a genuinely measured GFR, useful when an estimate would be misleading — before donating a kidney, before certain drug regimens, or when creatinine-based estimates disagree with the clinical picture.
Routine practice instead estimates GFR from an endogenous marker. Creatinine is the usual choice: it is produced at a fairly steady rate from muscle creatine, released into plasma, filtered at the glomerulus, and also secreted in small amounts by the proximal tubule. Because of that tubular secretion, creatinine clearance overestimates true GFR by roughly 10% to 20% in people with normal function.
The relationship between plasma creatinine and GFR is inverse and nonlinear. Doubling creatinine roughly halves GFR, but at low creatinine values the curve flattens, so small changes near the normal range mean little while the same absolute change at a higher level is important. This is why the estimating equations use log-transformed creatinine rather than a simple ratio.
Creatinine also has blind spots that no equation can fix. Very low muscle mass — frailty, amputation, advanced liver disease — produces less creatinine and therefore a falsely reassuring estimate. Very high muscle mass, creatine supplementation, or a recent cooked-meat meal pushes the estimate in the other direction. Drugs such as trimethoprim and cimetidine block tubular creatinine secretion and raise plasma creatinine without any change in filtration.
Cystatin C avoids most of the muscle dependence. It is produced by all nucleated cells at a stable rate, filtered freely, and not secreted. Estimates based on cystatin C are more accurate in people at the extremes of muscle mass and in those whose creatinine result seems inconsistent with their clinical state. It is not perfect either: thyroid disease, obesity, inflammation, and corticosteroid use all shift it independently of kidney function.
When precision matters, the creatinine-cystatin C combination equation is preferred, because the two markers' errors are partly independent. Guidelines now recommend confirming an unexplained creatinine-based estimate with a cystatin C measurement rather than proceeding straight to a measured clearance study.
The Estimating Equations
Several equations have been published, and the differences between them matter. The calculator on this page reports MDRD, the 2009 CKD-EPI equation, a Mayo quadratic formula, and the bedside Schwartz formula for children. Current guidance from kidney organizations favors the 2021 CKD-EPI equation, which is described below so the older results can be interpreted in context.
2021 CKD-EPI creatinine equation
This is the equation most laboratories in the United States now use. It has no race coefficient, and the age term is applied as a power of 0.9938:
eGFR = 142 × min(SCr ÷ κ, 1)α × max(SCr ÷ κ, 1)−1.200 × 0.9938age × 1.012 if female
The constants are κ = 0.7 and α = −0.241 for women, and κ = 0.9 and α = −0.302 for men. The min and max terms split the curve at the reference creatinine, applying a steeper slope below it and a shallower one above.
A worked example: a 60-year-old woman with a serum creatinine of 1.0 mg/dL. Her ratio is 1.0 ÷ 0.7 = 1.4286, which is above 1, so the max term applies with the exponent −1.200 and gives 0.652. The age factor is 0.993860 = 0.689. Multiplying 142 × 0.652 × 0.689 × 1.012 gives about 64 mL/min/1.73 m².
2021 CKD-EPI creatinine-cystatin C equation
When both markers are available, a combined equation is more accurate and is recommended for confirmation:
eGFR = 135 × min(SCr ÷ κ, 1)α × max(SCr ÷ κ, 1)−0.544 × min(SCys ÷ 0.8, 1)−0.323 × max(SCys ÷ 0.8, 1)−0.778 × 0.9961age × 0.963 if female
Here α is −0.219 for women and −0.144 for men, with κ unchanged. Cystatin C is entered in mg/L, and the reference point is 0.8 mg/L for both sexes.
2009 CKD-EPI equation
The earlier version used four separate equations depending on sex and on a Black race term, with coefficients of 166, 163, 144, and 141. For a non-Black woman the equation is 144 × (SCr ÷ 0.7)−0.329 or −1.209 × 0.993age, with the exponent chosen by whether creatinine is below or above 0.7. The same 60-year-old woman with a creatinine of 1.0 receives an estimate of about 61 from this version, against 64 from the 2021 equation.
The race coefficient was removed because race is a social category rather than a biological variable, and because the adjustment assigned a higher estimated GFR to everyone recorded as Black regardless of their actual body composition. Removing it slightly lowered estimates for some people and raised them for others, with the largest effect on Black patients whose true GFR had been overestimated. A race-free equation applied to everyone is both more equitable and, in validation studies, at least as accurate overall.
MDRD and Mayo quadratic
The MDRD equation, published in 1999 and re-expressed for IDMS-traceable creatinine, is 175 × SCr−1.154 × age−0.203 × 0.742 if female × 1.212 if Black. It was the standard for a decade and remains in use in some laboratories and older reports. It underestimates GFR in people with normal or near-normal function, which is why it was largely replaced.
The Mayo quadratic formula was designed to address that underestimation, clamping creatinine at a minimum of 0.8 mg/dL and applying an exponential expression with terms for creatinine, its square, age, and sex. It performs reasonably in the normal range but is not used in routine practice.
Children
The bedside Schwartz equation is 0.413 × height in centimetres ÷ serum creatinine in mg/dL. It applies to children and adolescents, and the constant reflects the relationship between body length and muscle mass during growth. It should not be used in adults, and it should not be used in newborns, where different constants apply.
What Moves the Number
An estimated GFR is a snapshot influenced by anything that changes creatinine production, creatinine handling, or actual filtration. Some of these factors are transient and some are permanent, and separating them is the clinician's job rather than the calculator's.
Muscle mass. The largest non-renal influence on creatinine. Bodybuilders, people taking creatine supplements, and anyone with unusually high muscle mass will show a higher creatinine and a lower estimated GFR than their kidneys warrant. Frail older adults, people with amputations, and those with advanced liver disease show the opposite.
Food and hydration. A cooked meat meal raises creatinine for several hours. Severe dehydration reduces filtration temporarily. Neither effect is large enough to explain a major change on its own, but both matter at the margin.
Exercise. Strenuous activity, particularly resistance training, can raise creatinine transiently through muscle breakdown. A test taken the morning after a hard session may read lower than the true baseline.
Drugs that block tubular secretion. Trimethoprim, cimetidine, and some antiretroviral agents reduce creatinine secretion without changing filtration, so plasma creatinine rises and the estimate falls even though the kidney is fine. This is a laboratory artifact rather than kidney injury, and it is easy to misread.
Drugs that change filtration or damage the kidney. NSAIDs constrict the afferent arteriole and can reduce filtration, especially in combination with an ACE inhibitor or ARB and a diuretic. Aminoglycosides and amphotericin are directly toxic to tubules. Calcineurin inhibitors used after transplant cause afferent constriction. Lithium causes chronic tubulointerstitial damage. Contrast media can precipitate acute kidney injury, though the risk is concentrated in people with existing impairment.
Acute illness. Sepsis, heart failure, vomiting, diarrhoea, and surgery all reduce renal perfusion and depress GFR. An estimate taken during an acute admission describes that moment, not the patient's baseline.
Pregnancy. Plasma volume and renal blood flow rise early, and GFR increases by up to 50%. Creatinine falls to around 0.4 to 0.6 mg/dL. Standard equations were not validated in pregnancy and can mislead in either direction.
Thyroid disease and steroids. Both affect cystatin C independently of filtration, which is one reason a combined creatinine-cystatin C estimate is preferred when either condition is present.
Preparing for the Test
Most GFR estimates come from a standard blood draw that measures creatinine, often as part of a metabolic panel. The preparation is modest, and the goal is simply to avoid a result that reflects yesterday's dinner or yesterday's workout rather than usual kidney function.
Fasting. A fast of 8 to 12 hours is often requested, mainly because glucose and lipids are usually measured at the same time. Water is encouraged: dehydration lowers filtration and raises creatinine.
Meat and protein. Avoid a large cooked-meat meal in the 12 hours before the draw. Dietary creatinine is absorbed and temporarily raises plasma levels. A normal meal is not a problem; a steak dinner the night before can be.
Exercise. Skip strenuous training for about 24 hours beforehand. Resistance work in particular can raise creatinine through muscle damage.
Creatine supplements. These raise plasma creatinine directly and can make an estimate look worse than the kidneys are. Mention them to the person ordering the test rather than stopping them on your own.
Medications. Do not stop prescribed medicines unless told to. Do tell the clinician about NSAIDs, trimethoprim, cimetidine, diuretics, ACE inhibitors, ARBs, and any supplements, because several of these affect either creatinine handling or filtration.
Acute illness. If you are vomiting, diarrhoeal, or feverish, the result will reflect dehydration and illness. Rescheduling is usually better than interpreting a value taken during an acute episode.
Timing and comparison. Kidney function has a mild circadian rhythm, and the difference is small compared with the factors above. Consistency matters more than the exact hour: tests taken at a similar time of day, under similar conditions, are easier to compare. Bring previous results to the appointment so the new number can be read as a trend.
Protecting Kidney Function
There is no lifestyle that guarantees stable filtration, but a small number of measures have consistent evidence behind them, and most of them are also good for the heart and blood vessels. The kidneys are vascular organs, and the same risk factors that damage arteries damage them.
Control blood pressure. Hypertension is the second leading cause of kidney failure after diabetes, and treating it slows progression. Guidelines for people with chronic kidney disease generally target a systolic pressure below 120 mmHg when measured under standardized conditions and tolerated without symptoms. The exact target should be set by the treating clinician, since it depends on age, frailty, and other conditions.
Control blood glucose. Diabetes is the leading cause of kidney failure in most countries. Keeping HbA1c within an individualized target reduces the risk of albuminuria and progression, and the benefit persists for years after control is established.
Take the medications that have evidence. ACE inhibitors and ARBs reduce albuminuria and slow progression in people with reduced filtration and albuminuria. SGLT2 inhibitors reduce the risk of progression and cardiovascular events in a broad range of chronic kidney disease, including in people without diabetes. Finerenone reduces progression in diabetic kidney disease with albuminuria. These are decisions for a clinician, but they are worth asking about.
Avoid nephrotoxins. Chronic NSAID use is the most common avoidable exposure. Herbal and supplement products are poorly regulated and some are directly toxic. The combination of an NSAID with an ACE inhibitor or ARB and a diuretic is a well-recognized cause of acute kidney injury and should be avoided where possible.
Eat with the kidneys in mind. Limiting sodium to under 2 g a day — about 5 g of salt — lowers blood pressure and reduces albuminuria. A diet built around vegetables, fruit, whole grains, legumes, and lean protein is appropriate for most people. Potassium and phosphorus restriction is only needed in advanced disease and should be guided by blood results and a dietitian rather than applied broadly.
Protein intake. Around 0.8 g per kilogram of body weight per day is the usual recommendation for adults with chronic kidney disease who are not on dialysis. Higher intake is not automatically harmful, but very high intake offers no kidney benefit and makes the diet harder to manage.
Move, do not smoke, and limit alcohol. Around 150 minutes a week of moderate activity helps control blood pressure and weight. Smoking accelerates vascular damage, including in the kidney. Alcohol has no protective role, and heavy use is directly harmful.
Get the tests. Blood pressure, blood glucose or HbA1c, creatinine, and a urine albumin-to-creatinine ratio are the four measurements that detect most kidney disease early. The urine test is the one most often missed, and it is the one that finds disease while filtration is still normal.
Common Misconceptions
Kidney disease attracts more than its share of folklore, partly because it is silent for so long. These are the beliefs that most often cause unnecessary worry or, worse, false reassurance.
"One low result means kidney disease." It does not. Filtration falls temporarily with dehydration, illness, a heavy protein meal, and some drugs. Chronic kidney disease requires a persistent abnormality for three months, usually confirmed on repeat testing.
"Normal creatinine means normal kidneys." Creatinine can stay within the reference range until a substantial amount of function is gone, because the relationship is nonlinear. It also depends on muscle mass, so a frail person can have a "normal" creatinine with reduced filtration. The estimate is more informative than the raw value, and albuminuria is more informative still in early disease.
"Kidney disease always has symptoms." Early and moderate disease is usually silent. Fatigue, swelling, itching, nausea, and changes in urination tend to appear when filtration is already well below 30. That silence is the argument for testing people with risk factors rather than waiting for symptoms.
"Drinking more water protects the kidneys." Adequate hydration is sensible, and dehydration is harmful. Beyond adequacy, there is no good evidence that forcing extra fluid raises filtration or prevents kidney disease, and in advanced kidney disease or heart failure excess fluid is dangerous. Drink to thirst unless a clinician has advised otherwise.
"A high GFR means the kidneys are fine." A normal filtration rate with persistent albuminuria or blood in the urine still indicates kidney disease. Filtration and barrier integrity are separate measurements, which is why both belong in a basic assessment.
"Kidney disease only affects older people." Age is a risk factor, not a requirement. Glomerulonephritis, lupus, polycystic kidney disease, and congenital abnormalities can cause kidney disease in children and young adults.
"Dialysis is the only outcome." Most people with chronic kidney disease never reach kidney failure. Early detection, blood pressure control, and the medications described above change the trajectory for many patients, and the majority of those diagnosed at stage 3 die of cardiovascular causes long before dialysis would be needed.
"Creatine supplements destroy kidneys." In people with normal kidney function, creatine supplementation raises plasma creatinine and can make an estimate look worse, but there is no consistent evidence that it causes kidney damage at usual doses. Anyone with existing kidney disease should discuss it with their clinician.
When to See a Nephrologist
A nephrologist is a physician who specializes in kidney disease, hypertension, dialysis, and transplantation. Not everyone with an abnormal result needs one, and referral criteria are designed to direct specialist time to the people most likely to benefit.
Filtration below 30. This is stage G4 or G5, and it warrants specialist involvement because preparation for kidney replacement therapy, management of anaemia and bone disease, and medication review all become relevant.
Significant albuminuria. An albumin-to-creatinine ratio above 300 mg/g, or persistent moderate albuminuria with an uncertain cause, justifies assessment. Albuminuria at that level often indicates a specific glomerular disease that may be treatable.
Rapid decline. A loss of more than 5 mL/min/1.73 m² in a year, or a sustained drop of more than 10% in filtration with other evidence of disease, suggests an active process rather than age-related change. Rapid decline is one of the strongest predictors of eventual kidney failure.
Blood or protein in the urine without explanation. Persistent haematuria, especially with casts or protein, needs evaluation for glomerulonephritis and structural disease.
Difficult blood pressure. Hypertension that persists despite three agents, including a diuretic, is both a cause and a consequence of kidney disease and often requires specialist management.
Hereditary disease. A family history of polycystic kidney disease, Alport syndrome, or other inherited kidney disorders is a reason for early assessment, since screening and family planning decisions may depend on it.
Recurrent stones or complicated anatomy. Repeated stone formation, structural abnormalities, or recurrent urinary infections with kidney involvement are within the specialist's remit.
Electrolyte and acid-base problems. Persistent hyperkalaemia, metabolic acidosis, or abnormalities of sodium and water balance that are difficult to manage in primary care.
Before certain treatments. Some chemotherapy agents, contrast studies, and surgeries carry higher risk in people with reduced filtration, and a nephrologist may be asked to advise on dosing and prevention.
Referral is not a sign that something has gone wrong with your care. It is the appropriate use of a specialist for a problem that benefits from specialist judgment, and earlier referral is generally associated with better preparation and fewer emergency starts on dialysis.
