1. Previous evidence suggests that the dose-dependent bioavailability of racemic propranolol may be partly due to product inhibition. We have examined this further by studying the individual enantiomers of propranolol in the perfused rat liver (IPRL) and in rat liver microsomes. 2. In recirculating IPRL experiments, (R)-propranolol (n = 7) or (S)-propranolol (n = 4) were infused at rates of 75, 150 and 231 nmol/min for three sequential 36-min phases. In single-pass experiments, (R)-propranolol (n = 4) or (S)-propranolol (n = 4) were administered at rates of 80, 136 and 239 nmol/min for three sequential 30-min phases. Steady-state bioavailability increased 10-20-fold over this dose range with both enantiomers in both recirculating and single-pass experiments. At the higher administration rates of (S)-propranolol, bioavailability in recirculating experiments was significantly greater than that in single-pass experiments, whereas there was no significant difference for (R)-propranolol. This suggests product inhibition of (S)- but not (R)-propranolol metabolism. 3. Of the metabolites examined, racemic 4'-hydroxypropranolol (4-OHP) inhibited the formation of 4-OHP, 5'-hydroxypropranolol (5-OHP) and desisopropylpropranolol (DIP) from (S)-propranolol and (R)-propranolol in microsomal studies (IC50 20 microM). Tissue levels of (S)-4-OHP in recirculating experiments (28.0 microM) at the highest dose (239 nmol/ min) of (S)-propranolol were greater than its IC50 of 20 microM, suggesting that 4-OHP is the inhibiting metabolite in the intact liver. The absence of evidence for product inhibition with (R)-propranolol in perfused livers suggests that (S)-4-OHP inhibits 4-hydroxylation of each isomer but (R)-4-OHP does not. 4. We conclude that in the recirculating IPRL, product inhibition of propranolol metabolism is evident with the (S)-isomer, but not he (R)-isomer, and that the inhibiting metabolite is (S)-4-OHP.