We sought to validate the improvement by adjustment for body surface area (BSA) of the accuracy of the original Cockcroft-Gault equation to estimate glomerular filtration rate (GFR), in a prospective cross-sectional study of 269 European patients with chronic kidney disease (CKD). We compared 3 methods: original Cockcroft-Gault equation, modified Cockcroft-Gault formula adjusted for BSA and abbreviated Modification of Diet in Renal Disease (MDRD) equation, using inulin clearance. Statistical analyses comprised repeated-measures analysis of variance (ANOVA), determination of the Pearson coefficient of correlation and a Bland-Altman concordance study. The ability of the GFR estimates to properly categorize patients in K/DOQI stages of CKD was also examined. Inulin clearance differed significantly from the standard Cockcroft-Gault method (ANOVA, p<0.001) and the abbreviated MDRD method (ANOVA, p<0.001) but not from the BSA-modified Cockcroft-Gault formula. Inulin clearance correlated better with the BSA-modified Cockcroft-Gault formula (r=0.88) and abbreviated MDRD equation (r=0.87) than with the standard Cockcroft-Gault equation (r=0.82). In concordance studies, bias was far smaller with the BSA-modified Cockcroft-Gault formula (mean bias -1.75 ml/min), than with the standard Cockcroft-Gault equation (mean bias -4.72 ml/min). The bias of the abbreviated MDRD was larger (mean bias +6.24 ml/min). Only patients with CKD stage 1 were better categorized with the BSA-modified Cockcroft-Gault formula and with the standard Cockcroft-Gault estimate than with the abbreviated MDRD equation. We conclude that adjustment for body surface area improves accuracy of the original Cockcroft-Gault equation.