This study determined whether sodium nitroprusside (SNP) changes cerebral vascular resistance during stable, hypothermic cardiopulmonary bypass (CPB). Cerebral blood flow (CBF) was measured using Xenon clearance in 39 patients anesthetized with fentanyl. In 25 patients (group 1), CBF was measured before and during infusion of SNP at a rate sufficient to reduce mean arterial pressure (MAP) approximately 20%. In 14 other patients (group 2), CBF was measured before and during simultaneous infusion of SNP and phenylephrine; SNP was continued at a rate that had reduced MAP approximately 20% while phenylephrine was added in a dose sufficient to restore MAP to preinfusion levels. Patients within each group were randomized to maintenance of PaCO2 approximately 40 mmHg (groups 1a and 2a), uncorrected for body temperature, or to maintenance of PaCO2 approximately 50 mmHg (groups 1b and 2b). The following variables were maintained within a narrow range: nasopharyngeal temperature (26-29 degrees C), pump oxygenator flow (1.7-2.5 l.min-1.m-2), PaO2 (150-300 mmHg), and Hct (22-28 vol%). In each patient, controlled variables varied no more than +/- 5% between measurements. In group 1a (PaCO2 approximately 40 mmHg), MAP was 86 +/- 9 mmHg (mean +/- SD) before and 65 +/- 8 mmHg during SNP infusion (P less than 0.0001). CBF was 12 +/- 3 ml.100g-1.min-1 before and 10 +/- 2 ml.100(-1).min-1 during SNP infusion (P less than 0.01). In group 1b (PaCO2 approximately 55 mmHg), MAP was 86 +/- 11 mmHg before and 66 +/- 13 mmHg during SNP infusion (P less than 0.0001).(ABSTRACT TRUNCATED AT 250 WORDS)