Abstract The purpose of these studies was to determine the effect of altered regional contractility on the linearity of regional left ventricular end-systolic relations. Significant change in the shape of these relations would limit their application as load-independent indices of regional contractility. In a paced, open-chest pig heart preparation ( n = 7), the left ventricular end-systolic pressure-segment length relation (ESPLR) and pressure-wall thickness relation (ESPTR) were obtained over a wide range of end-systolic pressures (134 ± 9 to 70 ± 6 mm Hg). Regional inotropic state was varied with intracoronary calcium and verapamil. The shapes of the ESPLR and ESPTR were characterized by using linear and quadratic models. Both provided a good fit, although the quadratic model showed a slight concavity to the segment length and thickness axes (second-order coefficient <0). In the linear model, calcium increased the slope of the ESPLR by 111% ( p < 0.01) and the slope of the ESPTR by 170% ( p < 0.01). At a pressure of 100 mm Hg, end-systolic segment length (L 100) shifted to the left ( p < 0.05) and end-systolic wall thickness (T 100) to the right ( p < 0.025). Verapamil decreased the slope of the ESPLR by 45% ( p < 0.01) and of the ESPTR by 33% (p not significant) and produced significant shifts in L 100 ( p < 0.001) and T 100 ( p < 0.025). The values of L 100 and T 100 determined by the quadratic fit were nearly identical to those for the linear fit, and both showed similar significant shifts with altered contractility. There was no significant change in the shape of the quadratic fit (as assessed by the second-order coefficient) with different contractile states. It is concluded that the curvilinearity of the ESPLR and ESPTR under physiologic conditions is slight and appears to be independent of the contractile state. Furthermore, a linear model of regional end-systolic relations can be used to assess regional left ventricular function in intact hearts.