With the single-pass, isolated, perfused rat heart preparation, we examined the effect of different perfusate albumin concentrations on the capillary permeability and pharmacodynamic effects of quinidine. Nine hearts were perfused for five consecutive 35-min phases with buffer containing quinidine and albumin concentrations of 0%, 0.1%, 1%, 6% and 0%, in that order, with a 55-min washout between each phase. Compared with 0% albumin perfusate, the equilibrium rate of quinidine perfusate output concentration (Cout) was slower with 0.1% albumin but faster with 1% and 6% albumin. Quinidine Cout, unbound fraction (fu) and perfusion flow rate (Q) for each phase of each experiment were fitted by a modified Kety-Renkin-Crone equation. Estimates of the capillary permeability-surface product (PS) for the two 0% albumin phases (14.4 +/- 5.6 and 12.3 +/- 4.4 ml/min) were significantly higher than those for the three albumin phases (5.3 +/- 1.0, 6.7 +/- 1.2 and 7.4 +/- 2.2 ml/min, respectively; P < .05). There was a direct, linear relationship between lengthening of the QT interval of the ECG and total and unbound quinidine concentrations, but the relationship for unbound concentration was independent of fu, showing that the pharmacodynamic effect was mediated by unbound concentration. In four additional experiments with antipyrine instead of quinidine, PS for the two 0% albumin phases (24.7 +/- 6.5 and 23.2 +/- 2.4 ml/min) was not significantly different than that for the three albumin phases (25.5 +/- 6.1, 22.5 +/- 2.9 and 22.9 +/- 3.6 ml/min, respectively). Albumin had no effect on the volume of distribution referenced to unbound drug for either drug.(ABSTRACT TRUNCATED AT 250 WORDS)