We studied the effect of ejection on end-systolic pressure in isolated heart preparations. Ejecting beats were compared with isovolumic beats having the same volume as at end systole. While holding end-systolic volume constant, various stroke volumes, including negative stroke volumes (volume injected during systole), were imposed using a predetermined volume command. After switching contraction mode between ejecting and isovolumic, we measured the immediate and steady changes in end-systolic pressure. In the first isovolumic beat after switching from steady-state ejecting beats, the change in end-systolic pressure was variable, depending on the stroke volume. The end-systolic pressure of the ejecting beat exceeded that of the isovolumic beat on average by up to 18 mm Hg with small stroke volume, but the ejecting end-systolic pressure became lower than isovolumic with either large stroke volume (stroke volume/end-systolic volume <0.96) or with negative stroke volume. During the transient phase following a switch from ejecting to isovolumic, the end-systolic pressure gradually decreased to a steady state. Consequently, even in steady state, ejecting end-systolic pressure exceeded isovolumic pressure over a significant range of stroke volume (stroke volume/end-systolic volume <1.18). After returning contraction mode from isovolumic back to ejecting, we observed responses that were a mirror image. These results indicated that in addition to negative uncoupling effect, ejection exerts positive effects on ventricular end-systolic pressure that are manifest both quickly and gradually. We hypothesized that the mechanism responsible for the positive effect is length-dependent activation via the larger volume (both at the initiation of contraction and averaged over a cardiac cycle) of a beat that ejects compared to one held isovolumic at end-systolic volume. The results with volume injection were consonant with this concept.