Both systolic and diastolic dysfunction have been described during pacing-induced ischemia, but the temporal sequence of systolic and diastolic impairment has not been established. Accordingly, 22 patients with coronary artery disease were paced at increasing heart rates and studied with simultaneous hemodynamic monitoring, electrocardiographic recording, and radionuclide ventriculography. In addition, with synchronized left ventricular pressure tracings and radionuclide volume curves, three sequential pressure-volume diagrams were constructed for each patient corresponding to baseline, intermediate, and maximum pacing levels. Eleven patients (group I) demonstrated a nonischemic response to pacing tachycardia without chest pain, significant electrocardiographic changes, or significant rise in left ventricular end-diastolic pressure (LVEDP) in the immediate postpacing period. These patients demonstrated a progressive decrease in LVEDP, end-diastolic volume, and end-systolic volume, no change in cardiac output or left ventricular ejection fraction, and a progressive increase in left ventricular diastolic peak filling rate and the end-systolic pressure-volume ratio. Pressure-volume diagrams shifted progressively leftward and slightly downward, suggesting both an increase in contractility and a mild increase in left ventricular distensibility. The remaining 11 patients (group II) exhibited an ischemic response to pacing tachycardia, with each patient experiencing angina pectoris, demonstrating greater than 1 mm ST segment depression on the electrocardiogram, and exhibiting greater than 5 mm Hg rise in LVEDP immediately after pacing. LVEDP, end-diastolic volume, and end-systolic volume in these patients initially decreased and then subsequently increased during angina, with no change in cardiac output but a decrease in ejection fraction. Left ventricular peak diastolic filling rate and the left ventricular end-systolic pressure-volume ratio both increased at the intermediate pacing rate but fell at maximum pacing. Pressure-volume diagrams for these patients shifted leftward initially, then back to the right, during intermediate and peak pacing levels, often with an upward shift in the diastolic pressure-volume relationship. LVEDP in group II was significantly higher than that in group I at the intermediate pacing level with no difference in end-diastolic or end-systolic volumes, suggesting decreased left ventricular distensibility in these patients before the onset of systolic dysfunction at the maximum pacing level. We conclude that an ischemic response to pacing tachycardia involves both systolic and diastolic dysfunction, with diastolic impairment often preceding systolic depression.