The respective roles of load level and loading sequence of the left ventricle (LV) are controversial in the in situ heart. They were analyzed under autonomic blockade and sedation in 17 dogs previously instrumented with a pressure micromanometer and ultrasonic crystals measuring LV diameters and wall thickness for computation of LV volume and stress. The time constant of isovolumic pressure fall (T) and end-systolic pressure (ESP) were calculated during the control state, caval occlusion, aortic constriction obtained by inflation of a hydraulic cuff occluder positioned around the aorta, and during the inflation of an intra-aortic balloon. Caval occlusion significantly decreased both ESP (from 124.0 +/- 6.6 to 88.7 +/- 3.7 mmHg; P less than 0.005) and T (from 29.0 +/- 2.2 to 18.8 +/- 2.4 ms; P less than 0.005), which were linearly correlated (mean r = 0.90 +/- 0.03) and inflation of an intra-aortic balloon increased both ESP (from 107.3 +/- 7.1 to 150.6 +/- 10.4 mmHg; P less than 0.005) and T (from 24.6 +/- 2.1 to 32.7 +/- 2.3 ms; P less than 0.005). Both interventions did not modify the loading sequence (analyzed by the evolution of systolic wall stress vs. time). In contrast, aortic constriction delayed to midsystole the time to which wall stress reached its peak and, for matched ESP with intra-aortic balloon inflation, T was not significantly different from control. We conclude that both the level of afterload and the loading sequence of LV are the determinants of T when contractility is not modified.