Neonates undergoing cardiac surgery at The Children's Hospital of Philadelphia frequently developed metabolic acidemia as they passively cooled prior to the start of cardiopulmonary bypass. This study was performed in an attempt to identify the mechanism for this acidemia. After receiving an initial dose of fentanyl (10 micrograms/kg) and pancuronium, 22 neonates were randomly assigned to maintain normothermia by active warming (Group I), or to permit passive cooling (Group II) before surgery. Arterial blood samples were obtained prior to, and at 10 and 45 min after entering the operating room for the analysis of pH, gas tensions, lactate, pyruvate, plasma free fatty acids, acetoacetate, beta-hydroxybutyrate, total CO2, and glucose concentrations. In the last 11 patients studied, the observation period was extended to 75 min at which time another arterial blood sample was obtained. There was a steady decrease in heart rate as the Group II patients cooled; however, arterial blood pressure did not change in either group. There were no changes in blood values measured in Group I neonates. In the Group II patients, there was a progressive decline in calculated base excess, total CO2, and an increase in serum lactate as the patients cooled. The metabolic acidemia that develops in neonates represents lactate accumulation as a consequence of surface cooling prior to surgery and the institution of cardiopulmonary bypass. Whether lactate accumulates as a result of anaerobic metabolism in underperfused tissue beds or reduced hepatic clearance could not be distinguished in this study. Since neither clinically significant hemodynamic changes nor differences in outcome were found between the two groups, the authors believe this mild lactic acidemia is inconsequential and does not require therapy.