In isovolumically beating hearts, the speed of left ventricular (LV) relaxation is uniquely determined by peak active stress (_max). In contrast, such a succinct description of relaxation is lacking for the ejection beats, although ejection is generally thought to hasten relaxation. We set out to determine how ejection modifies the relaxation-_max relationship obtained in the isovolumically beating hearts. Experiments were performed on five isolated rabbit hearts subjected to various loading conditions. Instantaneous LV pressure and volume were recorded and converted to active stress, from which isovolumic relaxation time (T_r) was defined as the time for stress to fall from 75 to 25% of _max (isovolumic beats) or its end-ejection value (ejection beats). Steady-state and transient isovolumic beat and steady-state ejection beat data were used to develop a multiple regression model. This model identified stress, current beat ejection, and previous beat ejection history as independent predictor variables of T_r and fit the data well in all hearts (r² > 0.98). Furthermore, this model could predict relaxation in transient ejection beats (r² = 0.80 for all hearts). Whereas the coefficient for the current beat ejection was negative (i.e., negative effect or hastening relaxation), the ejection history coefficient was positive (i.e., positive effect or slowing relaxation). The sum of these two coefficients was negative, corresponding to the commonly observed net negative effect of ejection on relaxation. The expected positive inotropic effect of ejection was also observed. The dissipations of both positive inotropic and relaxation effects were slow, suggesting a nonmechanical underlying mechanism(s). We postulate that these two effects are linked and caused by ejection-mediated changes in myofilament Ca²⁺ sensitivity.