Cardiopulmonary bypass is widely believed to be injurious to renal function. The low incidence of renal dysfunction with modern techniques of bypass led us to reexamine this concept by monitoring urine output and creatinine clearance in 18 adult patients undergoing nonpulsatile, hemodilution cardiopulmonary bypass for coronary artery bypass grafting (12 patients) or valve procedures (6 patients). Samples were taken before, during (mean duration of bypass, 105 +/- 26 minutes [+/- standard deviation]), and every two hours after bypass for 24 hours. Urine output (42 +/- 37.7 mL/h) and creatinine clearance (57 +/- 40.4 mL/min) were surprisingly low in the period before cardiopulmonary bypass (all values normalized to a body surface area of 1.73 m2). Urine volumes rose to 305 +/- 149.6 mL/h and creatinine clearance to 252 +/- 176.9 mL/min during bypass and decreased to stable values after eight hours in the postoperative unit (urine output, approximately 60 mL/h, and creatinine clearance, approximately 75 mL/min). Renal dysfunction did not develop in any patient. Nine patients who required loop diuretics for low urine output 18 hours postoperatively had a sustained increase in both urine output and creatinine clearance lasting up to six hours. We conclude the following: modern techniques of cardiopulmonary bypass are not injurious to renal function; urine output and creatinine clearance are decreased before cardiopulmonary bypass, probably because of preoperative dehydration; and loop diuretics in the postoperative period increase both urine output and creatinine clearance for as long as six hours after administration.