The effect of sepsis on the intrinsic contractile status of the myocardium is best examined in the awake, closed-chest animal with intact circulation because anesthesia, open thoracotomy, and circulatory support are all known to affect hemodynamics. To fulfill these criteria, 18 adult dogs were chronically studied in the awake state after instrumentation with left ventricular high-fidelity pressure catheters and ultrasonic dimension transducers to measure left ventricular transmural pressure and minor axis dimension. This allowed computer assessment of the left ventricular end-systolic pressure-dimension relationship in the control state and at intervals following cecal ligation in one group of dogs. A second group of control animals was studied over variable time intervals without cecal ligation to evaluate the temporal stability and reproducibility of the animal model and the end-systolic pressure-dimension relationship. Evaluation of contractility by use of the end-systolic pressure-dimension relationship was essential because this relationship is a sensitive indicator of the intrinsic myocardial contractile state while remaining insensitive to the wide swings in preload and after load that are commonly seen in sepsis. In the control group of dogs, the temporal consistency and stability of the end-systolic pressure-dimension relationship in this model was confirmed; no significant changes in the slope and dimension-axis intercept were demonstrated over the study interval. In the septic group of dogs, however, the intrinsic myocardial contractility significantly deteriorated as the mean slope of the end-systolic pressure-dimension relationship (mmHg/mm) decreased from 16.87 +/- 0.85 to 12.79 +/- 1.67 over 120 hours following cecal ligation. Intrinsic contractility of the heart during sepsis was therefore isolated for the first time from the widely variant loading conditions seen during sepsis by pressure-dimension analyses in the chronically instrumented, awake, closed-chest canine with intact circulation.