What creatinine clearance is, and why the 1976 equation survives
Creatinine clearance is the volume of plasma the kidneys clear of creatinine each minute. Creatinine is a useful marker because muscle produces it at a roughly constant rate and the glomerulus filters it freely. It is an imperfect marker because the proximal tubule also secretes a little of it, so clearance overstates true glomerular filtration by perhaps 10 to 20% in health and considerably more in advanced kidney disease.
Cockcroft and Gault published their estimate in 1976 from 249 timed collections in adult inpatients. Its structure encodes three facts. Creatinine production falls with age, so the numerator carries (140 − age). Production scales with muscle mass, for which body weight is the available proxy, so weight multiplies. Women carry less muscle for a given weight, so the result is multiplied by 0.85. The constant 72 makes the units come out in mL/min when creatinine is in mg/dL.
Newer equations estimate glomerular filtration rate more accurately, and for staging chronic kidney disease you should use one. Cockcroft-Gault survives because the pharmacokinetic studies that generated the renal dose adjustments in drug labelling were themselves analysed against Cockcroft-Gault clearance. When a label says to reduce a dose below 30 mL/min, that threshold was validated on this equation, in mL/min, unnormalised. Using a different estimate can put a patient in a different dosing band for no clinical reason.
Which weight to use — the part that actually changes the answer
Weight appears linearly in the numerator, so the weight you choose scales the result directly. In an obese patient the difference between total and ideal body weight can be 50 mL/min, which is the difference between full dose and half dose for several antibiotics and anticoagulants.
Cockcroft and Gault used measured weight. Their cohort, though, contained few obese patients, and adipose tissue produces very little creatinine while adding a great deal of weight. Using total body weight in obesity therefore inflates the estimate. Using ideal body weight alone under-counts, because obesity does add some lean mass and does increase filtration. The compromise that pharmacy practice settled on is adjusted body weight: ideal weight plus 40% of the excess over ideal.
The convention this calculator applies, and names on screen, is the one taught in most clinical pharmacy curricula:
- If measured weight is below ideal body weight, use measured weight. Substituting the larger ideal weight would raise the estimate and risk overdosing.
- If BMI is 30 kg/m² or more, use adjusted body weight.
- Otherwise use ideal body weight.
Ideal body weight comes from the Devine formula: 50 kg for men or 45.5 kg for women, plus 2.3 kg for every inch of height above 5 feet. It is an approximation built on imperial units, which is why this calculator shows the height conversion as an explicit step. None of this weight-selection logic is part of the original equation — it is convention layered on top, and different institutions use different variants. Follow your own protocol, and use this calculator to see how much the choice moves the answer.
Worked example: a 68-year-old man, 82 kg, 175 cm, creatinine 1.3 mg/dL
Work every step by hand.
- Height in inches. 175 ÷ 2.54 = 68.90 in, so 8.90 inches over five feet.
- Ideal body weight. 50 + 2.3 × 8.90 = 50 + 20.46 = 70.46 kg.
- BMI. 82 ÷ 1.75² = 82 ÷ 3.0625 = 26.8 kg/m². Below 30, and measured weight exceeds ideal, so ideal body weight is the dosing weight.
- Age term. 140 − 68 = 72.
- Denominator. 72 × 1.3 = 93.6.
- Clearance on ideal weight. 72 × 70.46 ÷ 93.6 = 5,073 ÷ 93.6 = 54.2 mL/min.
- Compare with total weight. 72 × 82 ÷ 93.6 = 63.1 mL/min — nearly 9 mL/min higher, and on the other side of the 60 mL/min threshold several labels use.
- Normalise for comparison with eGFR. Body surface area by Mosteller is √(175 × 82 ÷ 3600) = √3.986 = 1.997 m². Then 54.2 × 1.73 ÷ 1.997 = 47.0 mL/min/1.73 m².
The last two lines are the practical lesson. The same patient is 63, 54 or 47 depending on which convention you apply, and only one of those numbers belongs in a dosing decision. Dose from the absolute clearance in mL/min on the weight your protocol specifies; use the normalised figure only when you want to line the result up against a reported eGFR.
Reading the result, and when not to believe it
For drug dosing, read the absolute clearance in mL/min against the thresholds printed in the product label — commonly 60, 50, 30 and 15 mL/min. Do not convert, do not normalise, and do not substitute an eGFR in mL/min/1.73 m² unless the label says you may. In a small patient the normalised figure runs higher than the absolute one and in a large patient it runs lower, so the substitution errs in a different direction depending on body size.
The estimate assumes a steady state. In acute kidney injury the serum creatinine lags the true filtration rate by a day or more: a patient whose kidneys stopped this morning still has a near-normal creatinine this afternoon, and any equation fed that creatinine will report near-normal clearance. When creatinine is moving, no estimating equation is valid, and dosing must be based on the trajectory and clinical judgement.
Muscle mass is the other systematic problem. Amputation, paralysis, prolonged critical illness, cirrhosis and advanced cancer all lower creatinine production, so creatinine stays low while filtration falls, and the equation overestimates clearance — sometimes grossly. The mirror image occurs in a heavily muscled young adult or someone taking creatine supplements. Drugs that block tubular creatinine secretion without changing filtration, including trimethoprim, cimetidine and several tyrosine kinase inhibitors, raise creatinine by 10 to 30% within days and make clearance look worse than it is.
A timed urine collection sidesteps the estimating equation but introduces its own error: a missed void or a mistimed start makes the result meaningless. Check the total creatinine excreted in the collection against the expected 15 to 20 mg per kg per day for women and 20 to 25 mg per kg per day for men before you trust it. A collection well below that range was incomplete.
How the weight basis changes the estimate at the same creatinine
| Measured weight | BMI (kg/m²) | CrCl on total weight | CrCl on ideal weight | CrCl on adjusted weight |
|---|---|---|---|---|
| 55 kg | 17.4 | 61.1 | 81.1 | 73.1 |
| 73 kg | 23.1 | 81.1 | 81.1 | 81.1 |
| 90 kg | 28.5 | 100.0 | 81.1 | 88.7 |
| 120 kg | 38.0 | 133.3 | 81.1 | 102.0 |
| 150 kg | 47.5 | 166.7 | 81.1 | 115.3 |
At 73 kg the three bases coincide because measured weight equals ideal weight. The spread widens in both directions from there — 166.7 against 81.1 mL/min, a gap of 85.6, between total and ideal weight at 150 kg.
Common errors
- Feeding creatinine in µmol/L into a mg/dL formula. An 88-fold error. Divide µmol/L by 88.4, or use the unit selector above.
- Using total body weight in obesity without checking the protocol. It can double the estimate relative to ideal weight.
- Dosing from a normalised clearance. mL/min/1.73 m² is for comparing patients; mL/min is for dosing one.
- Using it in acute kidney injury. Creatinine lags filtration by a day or more, so the estimate is reassuringly wrong exactly when it matters.
- Applying it to children. The equation was derived in adults; paediatric estimation uses the Schwartz equation with height and a different constant.
- Ignoring drugs that block creatinine secretion. Trimethoprim and cimetidine raise creatinine without changing filtration.
- Believing an incomplete urine collection. Check the 24-hour creatinine excretion against the expected mg-per-kg range before using a measured clearance.
- Rounding creatinine up to 1.0 by reflex. It is a local convention with no validation behind it, and it lowers the estimate. Follow policy, but know what you are doing.
Cockcroft-Gault or CKD-EPI?
Use Cockcroft-Gault in mL/min for renal drug dosing, because that is the equation and the unit against which dose adjustments in product labelling were derived. Use CKD-EPI 2021 eGFR in mL/min/1.73 m² for diagnosing and staging chronic kidney disease, because it is more accurate and is the basis of the KDIGO stages. The two answer different questions and disagree routinely, particularly at the extremes of body size. If a dosing reference explicitly permits eGFR, de-normalise it first by multiplying by the patient's body surface area and dividing by 1.73.
Related bedside calculations
Renal dosing rarely stops at clearance. The dose itself usually comes from a weight-based or surface-area-based order, so the weight-based dosage calculator and the body surface area calculator are the natural next steps; the latter uses the same Mosteller formula this page uses to normalise clearance. If the drug arrives as a powder, the reconstitution calculator converts the ordered dose into a volume.
On the diagnostic side, when creatinine has risen acutely the question is usually prerenal versus intrinsic, which is what the fractional excretion of sodium addresses. If the same patient has a metabolic acidosis, the anion gap distinguishes uraemic acidosis from the alternatives, and severe uraemia commonly comes with the electrolyte disturbances that make corrected calcium worth checking.