The effects of mechanical changes in loading conditions on the left ventricular end-systolic pressure-volume relation (ESPVR) were studied in nine open-chest dogs, including three dogs studied before and after beta-adrenergic blockade. Left ventricular pressure was measured with a micromanometer, and left ventricular volume was measured with a conductance catheter. ESPVRs were obtained by increasing left atrial inflow over wide volume ranges (as much as threefold) under three different conditions: control or high or low aortic impedance. High impedance was obtained by occlusion of the descending aorta, and low impedance was obtained by a shunt between the subclavian artery and the left atrium. In the unblocked animals in 21 of 28 runs, a second-order polynomial equation gave a better fit for the ESPVR than a linear relation. To quantify the effects of the changes in aortic impedance on the ESPVR, we calculated from the quadratic equation its volume intercept (V18) and its local slope (E18) at an end-systolic pressure (Pes) of 18 kPa. In the unblocked animals, a statistically significant difference was found in V18 between low impedance (21.50 +/- 6.27 ml) and high impedance (14.10 +/- 8.98 ml; p less than 0.005) and between control (19.14 +/- 9.58 ml) and high impedance (p less than 0.05). In most dogs, E18 was increased at high and decreased at low impedance, but not significantly. In the additional experiments with beta-blockade, the nonlinearity diminished somewhat, but the load dependency of the ESPVR remained present after beta-blockade because the same leftward shift of the ESPVR with high aortic impedance was found. Two other relations, namely, of dP/dtmax and of stroke work versus end-diastolic volume, were also investigated, which on the whole showed the same behavior as the ESPVR. These results indicate that the ESPVR and dP/dtmax-Ved and stroke work-end-diastolic volume relations, when studied over a wide volume range, are nonlinear and that changes in loading conditions influence indexes of contractility derived from these relations, especially the volume intercepts, in such a way that an increase in aortic impedance may be interpreted as an increase in contractility. Blocking the beta-adrenergic receptors did not influence the load dependency of the ESPVR but, in some cases, tended to decrease the nonlinearity in concordance with the relation between contractility and nonlinearity in isolated hearts.