Timolol, like propranolol, is a nonselective beta-adrenergic blocker, but it is much less lipid soluble and is formulated as a single enantiomer rather than a racemic mixture. We examined the effects of such differences on bioavailability, systemic clearance, and pharmacologic response. Ten healthy subjects received placebo, 0.2 mg/kg IV propranolol, 3.2 mg/kg oral propranolol, 0.025 mg/kg IV timolol, and 0.4 mg/kg oral timolol in double-blind, randomized crossover fashion. Plasma concentrations of total drug were monitored up to 8 hr, while responses to submaximal exercise were measured at 0, 2, and 6 hr. Timolol had greater bioavailability (F = 0.61 +/- 0.06(SEM) and 0.32 +/- 0.04) and lower systemic clearance (463 +/- 74 ml kg-1 hr-1 and 1040 +/- 120 ml kg-1 hr-1) than propranolol. Half-lifes were of the same order (2.7 and 2.9 hr). There was a marginal correlation between areas under the intravenous and oral plasma concentration-time curves for timolol (r = 0.66, P = 0.054), but none for propranolol (r = 0.48, P NS). There were also correlations of the response (percent reduction in heart rate) to oral and intravenous timolol at 2 hr (r = 0.72, P less than 0.05) and 6 hr (r = 0.84, P less than 0.01) hr, but none between responses to oral and intravenous propranolol. Finally, the response to oral timolol was related to the area under its plasma concentration-time curve, whereas that to propranolol was not. We conclude that the physicochemical properties of timolol lead to greater bioavailability. Pharmacologic response to an oral dose of timolol, unlike that to propranolol, can be predicted from the response to an intravenous dose and reflects its plasma concentration.