In 128 consecutive operations, a prospective, randomized comparison of morphine and halothane as primary agents for cardiac valvular repair and/or replacement was made. During induction, average systolic and mean blood pressures were lower (9 torr and 6 torr, respectively) with halothane than with morphine. However, the incidences of serious hypotension with the two agents were similar. Six of 61 patients with halothane and seven of 67 patients with morphine reached systolic blood pressures of 70 torr or less before incision. Intraoperative hypertension was more frequent and more severe with morphine, necessitating frequent use of supplementary chlorpromazine. Cardiac output data suggest that the higher average blood pressure seen with morphine was related to increased peripheral vascular resistance rather than increased cardiac output. There appeared to be hemodynamic differences between the agents, but neither mortality rates nor durations of hospital stay or postoperative stay in the intensive care unit demonstrated a clear–cut advantage of either morphine or halothane for anesthesia during cardiac–valve operations.