Measured glomerular filtration rates (mGFRs) were obtained by (99)mTc-DPTA, (125)I-iothalamate, iohexol, (51)Cr-EDTA, non-radiolabeled iothalamate, or inulin clearance from centers agreeing to perform mGFR in six completed and one ongoing Wyeth Research multicenter trials evaluating sirolimus (SRL) in regimens with or without a calcineurin inhibitor (CNI). Estimated GFRs (eGFRs) were calculated by the Cockcroft-Gault (eGFR(CG)), Nankivell (eGFR(NK)), and simplified Modification of Diet in Renal Disease (eGFR(MDRD)) equations. Bias, precision, and accuracy for each of these equations were estimated by tertiles and by regimen. For the Rapamune Maintenance Regimen (RMR) trial, eGFR outcomes were also compared between treatments {[SRL-cyclosporine (CsA) versus SRL]} using the three eGFR formulas. In the lowest mGFR tertile (6-40 ml/min), eGFR(MDRD) gave the best accuracy with the least bias whereas eGFR(NK) and eGFR(CG) performed better in the highest mGFR tertile (58-139 ml/min). At 24 months in the RMR study, mean differences in eGFR between treatments were 13.6, 14.2, and 13.5 ml/min/1.73 m(2) for eGFR(CG), eGFR(NK), and eGFR(MDRD), respectively, favoring CsA withdrawal (P-values for all <0.001). The accuracy of the three eGFR equations was affected by mGFR range but not by immunosuppressive regimens utilizing SRL, SRL-CNI or CNI-based therapy.