The reversible upward shift of the diastolic pressure-volume curve that occurs during pacing-induced ischemia has not been fully explained by increases in passive chamber stiffness or reductions in relaxation rate. We measured the fully relaxed pressure-volume relation defined by both filling and nonfilling beats and the isovolumic relaxation time constant in nonfilling beats before and during demand ischemia using our in situ left ventricular volume clamping technique in 10 dogs. Pacing-induced ischemia shifted the diastolic pressure-volume curves in filling beats upward compared with the end-diastolic pressure-volume relation of the normally perfused heart. In contrast, the end-diastolic points for nonfilling beats during pacing-induced ischemia fell on the fully relaxed pressure-volume relation defined by the normally perfused heart. Left ventricular filling per se was necessary for the upward shift of the diastolic pressure-volume curve observed during pacing-induced ischemia. We speculate that active force developed during diastole induced by stretch activation or, perhaps, length-dependent changes in calcium sensitivity of the myofilaments in the ischemic myocardium due to stretch of the myocardium during rapid diastolic filling may contribute to the upward shift of the diastolic pressure-volume curve observed during pacing-induced ischemia.