The end-systolic pressure-volume relationship (ESPVR) is proposed and used as a reliable index of left ventricular (LV) contractility despite the fact that its afterload independence has been challenged. Furthermore, the physiological relevance of its volume-axis intercept, V(0), remains unclear. Systemic haemodynamics and pressure-volume loops obtained by inferior vena cava occlusion were recorded in 21 rats anaesthetized by isoflurane inhalation and instrumented with a conductance pressure-volume catheter in response to incremental I.V. doses of adrenaline, dobutamine, phenylephrine, metoprolol, papaverine and isoflurane inhalation. In conditions with large variations (± 100%) of both inotropy and afterload, infusion of negative inotropic drugs was associated with a dose-dependent rightward shift of ESPVR accompanied by a decrease in its slope (end-systolic elastance, E(es)), whereas positive inotropic agents produced an isolated decrease in V(0). With the predominant vasoactive drugs, there was a dose-dependent change in E(es) without major horizontal shifts, demonstrating that this slope mainly represents LV afterload rather than inotropy. When contractility was altered, V(0) was negatively correlated to the preload-adjusted contractility index, PAdP/dt(max), demonstrating that a reduced V(0) provides a good reflection of increased LV contractility. From these results, we computed a logarithmically adjusted E(es)/V(0) ratio, which resulted in reasonably strong concordance with PAdP/dt(max), including all the investigated drugs and dosages [n = 288; bias, 0.8 ± 16.2% (SD)]. Concordance with E(es) (bias, 7.2 ± 58.7%) or V(0) (bias, -0.6 ± 33.4%), used alone or with other commonly used contractility indices, was far less significant. In contrast to E(es), V(0) provides a relatively good LV contractility index because it is much less sensitive to afterload.