The effect of CPB on plasma ADH levels, urine flow, and urine osmolality was studied in nine patients. All patients received morphine, 1 mg. per kilogram, and 50 per cent nitrous oxide-50 per cent oxygen for anesthesia. CPB utilized a Travenol disposable bubble oxygenator and the prime consisted of 3 L. of Ringer's lactate. Measurements were made prior to induction of anesthesia , at 30 minutes following surgical incision, and at 15, 30, and 45 minutes during CPB. There were no statistically significant changes in mean arterial BP, cardiac index, serum sodium, or serum osmolality in any period. Urine flow increased from 0.99 +/- 0.3 ml. per minute to a high of 6.13 +/- 2.0 ml. per minute at 30 minute at 30 minutes on CPB (P less than 0.02). Urine osmolality declined from a control value of 691 +/- 142 mOsm. per kilogram to a low of 425 +/- 48 mOsm. per kilogram at 45 minutes on CPB (p less than 0.05). ADH levels rose from a control value of 4.3 +/- 1.5 to 13.0 +/- 3.3 pg. per milliliter with surgical stimulatiion (p less than 0.05). During CPB the ADH levels rose to a peak of 23.7 +/- 3.6 pg. per milliliter at 30 minutes (p less than 0.01) and were declining at 45 minutes. These data suggest that the stress of CPB results in an outpouring of ADH (or vasopressin) to function as a pressor to produce an increase in peripheral resistance. The ADH concentrations far exceed those required for normal physiologic control of water excretion and the urineflow will thus vary more with the hemodynamic changes than with the ADH levels.