The role of perfusion pressure and flow during cardiopulmonary bypass with moderate hypothermia and hemodilution in the development of new postoperative renal or clinically apparent cerebral dysfunction was examined in 504 adults. Cardiopulmonary bypass flow was targeted at greater than 40 mL.kg-1.min-1 and pressure at greater than 50 mm Hg. Flows and pressures less than target occurred in 21.6% and 97.1% of patients, respectively. Fifteen patients (3.0%) suffered new renal and 13 (2.6%) new central nervous system dysfunction. Low pressure or flow during cardiopulmonary bypass, expressed in absolute values or in intensity-duration units, were not predictors of either adverse outcome. Multivariate analysis identified use of postoperative intraaortic balloon counterpulsation (p less than 10(-6], excessive blood loss in the ICU (p less than 10(-4], need for vasopressors before cardiopulmonary bypass (p less than 10(-4], postoperative myocardial infarction (p less than 10(-3], emergency reoperation (p less than 0.002), excessive postoperative transfusion (p less than 0.02), and chronic renal disease (p less than 0.03) as independent predictors of postoperative renal dysfunction. Independent predictors of postoperative central nervous system dysfunction were cardiopulmonary resuscitation in the intensive care unit (p less than 10(-6], intracardiac thrombus or valve calcification (p less than 0.02), and chronic renal disease (p less than 0.03). Age greater than 65 years (40.7% of patients) did not predict either outcome. We conclude that failure of the native circulation during periods other than cardiopulmonary bypass rather than the flows and pressures considered here is the major cause of renal and clinically apparent central nervous system dysfunction after cardiac operations.