The time constant of left ventricular pressure fall, tau, has frequently been used as a measure of myocardial relaxation in the blood-perfused, ejecting heart. The aim of the present study was to characterise tau in relation to beta-adrenergic activation, coronary perfusion pressure and flow as well as cardiac oxygen supply and demand in the isolated, isovolumically beating heart. Therefore, tau was analysed from digitised left ventricular pressure data in a total of 23 guinea pig hearts perfused with saline at constant pressure (60 cmH2O). The coronary venous adenosine concentration ([ADO]) served as an index of myocardial oxygenation. Isoprenaline (0.4-3.2 nmol l-1) decreased and propranolol (3-9 mumol l-1) increased tau dose-dependently (linear regression tau vs lg([isoprenaline]), r = 0.74; tau vs. lg([propranolol]), r = 0.66, both P < 0.05). During graded reductions in cardiac oxygen supply from 96.1 +/- 12.6 (SEM) to 44.4 +/- 4.4 microliters min-1 g-1, tau was prolonged from 61.5 +/- 12.7 to 109.9 +/- 22.6 ms while left ventricular developed pressure (LVDP) decreased from 90.7 +/- 7.2 to 40.7 +/- 5.1 mmHg. In parallel, [ADO] increased from 23.7 +/- 9.1 to 58.0 +/- 19.1 pmol ml-1 (P < 0.05). Increasing oxygen supply to 165.4 +/- 32.4 microliters min-1 g-1 augmented LVDP to 102.7 +/- 7.3 mmHg but did not change tau or [ADO]. There was a dual response of tau to changes in cardiac oxygen supply or demand. As long as oxygen supply and demand matched, tau remained constant. However, when the oxygen supply was less than 100 microliters min-1 g-1, left ventricular relaxation was prolonged in parallel to the reduction in oxygen supply. In addition, a close relationship was observed between [ADO] as an indicator of myocardial oxygenation and tau (Spearman correlation, r = 0.99, P < 0.005). We conclude that the time constant of left ventricular pressure fall, tau, sensitively reflects myocardial relaxation in the isolated, isovolumically beating guinea pig heart. Moreover, in this model left ventricular relaxation is not influenced by alterations in coronary perfusion pressure or flow as long as cardiac oxygen demand is matched by an adequate supply. Rather relaxation is strictly coupled to myocardial oxygenation as reflected by coronary venous adenosine concentrations.