Rate-pressure product (RPP) is a sensitive index of myocardial oxygen consumption (mVO2) in awake people. We wished to determine whether this relationship persisted under anesthesia and in the face of concurrent large changes in myocardial contractility and left ventricular filling pressures. In 16 patients scheduled for coronary artery bypass surgery, we inserted coronary sinus and Swan-Ganz catheters, and a central aortic catheter via the brachial artery, before induction of anesthesia with either morphine (2 mg/kg) or halothane, chosen in random order. We measured aortic, pulmonary, and venous pressures, cardiac output, systolic time intervals, and thermodilution coronary sinus flow. We calculated mVO2 as coronary sinus flow times myocardial arteriovenous oxygen content difference. We found significant correlations between mVO2 and heart rate (r = 0.57), systolic blood pressure (r = 0.52), the index delta /delta T (r = 0.53, and RPP (r = 0.78). Multiple regression of RPP and delta P/delta T against mVO2 increased their correlation (r = 0.86), while multiple regression of RPP and pulmonary wedge pressure against mVo2 did not significantly improve the correlation of RPP alone (r = 0.75). We conclude that hemodynamic changes anesthesia and surgery do not decrease the sensitivity of RPP as an index of mVO2.