To test the hypothesis that the rate of left ventricular relaxation can be dissociated from left ventricular stiffness, indexes of stiffness and relaxation were evaluated during mild hypoxia and hypoxia with isoproterenol or caffeine in isovolumically beating, isolated rat heart preparations. In group 1, the hearts were perfused with mild hypoxic buffer for 30 min. In group 2, 10 min of mild hypoxic perfusion was followed by 20 min of mild hypoxia plus isoproterenol (10(-6) M). In group 3, the hearts were perfused with mild hypoxic buffer for 10 min followed by 20 min of hypoxia plus caffeine (10(-3) M) perfusion. In group 1, left ventricular end-diastolic pressure (LVEDP) was elevated only slightly, and time constant of left ventricular relaxation T was also prolonged slightly. In group 2, LVEDP was higher, but T was much shorter than the corresponding values in group 1 (LVEDP, 18 +/- 4 mmHg; T, 31.4 +/- 3.4 ms). In group 3, T was more markedly prolonged, and LVEDP tended to be higher than in group 1. These results indicate that the determinants of left ventricular relaxation and stiffness are not identical and that in certain conditions relaxation can be separated from left ventricular stiffness.