This randomized, double-blind study was designed to evaluate the effect of prostacyclin (epoprostenol) on the incidence and severity of postoperative neuropsychologic dysfunction in patients undergoing coronary artery operation. Four days before operation and 1 week after operation, 100 patients having coronary artery bypass grafting underwent detailed neurologic and psychologic examinations and computed tomographic scans of the brain. The psychologic examination was repeated 2 months after operation. During cardiopulmonary bypass, all patients received 300 U/kg of heparin and then either buffer-diluent or prostacyclin (12.5 ng/kg/min from the time of heparinization until onset of cardiopulmonary bypass and 25 ng/kg/min during cardiopulmonary bypass). No deaths or major neurologic complications occurred in this series. Ninety-six patients completed the psychologic and neurologic evaluations 1 week after operation; 74 of these patients were evaluated psychologically 2 months after operation. Psychologic testing demonstrated similar declines in postoperative performance in both the prostacyclin-treated and the control groups; these changes were no longer present in either group 2 months after operation. Results of neurologic examinations and computed tomographic scans of the brain were unchanged. We conclude that the administration of prostacyclin during cardiopulmonary bypass in patients undergoing routine coronary artery operation has no effect on perioperative cognitive changes.