Twelve patients requiring temporary pacing following acute myocardial infarction (AMI) (10 heart block, 2 junctional bradycardia) had hemodynamic measurements taken with ventricular demand pacing at 80 ppm (VVI80), ventricular demand pacing at the atrial rate (VVIa), physiological pacing (DDD), and spontaneous (intrinsic) rhythm. VVI80 mode did not improve any hemodynamic parameter compared with spontaneous rhythm. VVIa mode improved diastolic and mean arterial pressures only. DDD mode improved most hemodynamic parameters compared with spontaneous rhythm (cardiac output by 29% [P less than 0.0001]; blood pressure: diastolic by 24% [P less than 0.01], systolic by 19% [P less than 0.01], mean by 21% [P less than 0.005]; pulmonary wedge pressure by 10% [P = 0.057] and right atrial pressure by 24% [P less than 0.005]) and also significantly improved some parameters compared with VVIa (cardiac output by 20% [P less than 0.001], systolic blood pressure by 11% [P less than 0.01] and right atrial pressure by 15% [P less than 0.01]). Physiological pacing is hemodynamically superior both to ventricular pacing and spontaneous rhythm for patients requiring temporary pacing following AMI. Ventricular pacing at 80 ppm has little hemodynamic advantage over spontaneous rhythm.