Aims Long-term treatment with β1-selective adrenergic antagonists gives rise to cross-sensitisation of cardiac β2-adrenoceptor responses, with no corresponding alteration in β1-adrenoceptor responses. We performed a prospective randomised double-blind placebo-controlled cross-over study of the effects of nonselective β-blockade with timolol on α-adrenergic and angiotensin II receptor mediated responses in normal subjects. We also wished to study the time course of β1- and β2-adrenergic responses after withdrawal of timolol. Methods Six healthy males received timolol 10 mg twice daily or placebo for 14 days. On day 11 of treatment, vascular α1-, α2- and angiotensin II receptor responses were assessed by measuring the blood pressure increases in response to intravenous phenylephrine, α-methylnoradrenaline and angiotensin amide respectively, following one dose of timolol 10 mg (to block the β-adrenergic effects of phenylephrine and α-methylnoradrenaline). Both systolic and diastolic blood pressure increased in response to each of these drugs, but these increases were not different on timolol treatment or placebo. Following cessation of treatment with timolol or placebo, β1- and β2-adrenoceptor mediated responses were assessed by measuring the heart rate responses to treadmill exercise and intravenous salbutamol infusion respectively. Half each of the subjects underwent this 2 days and 3 days respectively, after the end of treatment. Results Both exercise-induced and salbutamol-induced tachycardia were not different following placebo or 3 days following the end of timolol treatment. However, 2 days following timolol treatment, both were attenuated; the reduction in salbutamol-induced tachycardia was significant, whilst the reduction in exercise tachycardia did not reach statistical significance. We also measured metabolic responses to exercise and to salbutamol infusion. Exercise induced a rise in plasma potassium and noradrenaline. Salbutamol produced a fall in plasma potassium, a rise in plasma glucose and insulin and also a rise in plasma noradrenaline. All of these changes were not different following placebo or 3 days after the end of timolol treatment; by contrast, 2 days following timolol treatment, all were significantly attenuated, with the exception of the rise in plasma glucose. In addition, the rise in both plasma glucose and insulin in response to an oral load of 75 g glucose were not different post-placebo, 2 or 3 days post-timolol. Conclusions These results show that, following 14 days of nonselective β-adrenoceptor blockade with timolol, there is evidence of residual β-adrenoceptor blockade 2 days after drug withdrawal; this finding is in contrast with the known plasma profile of timolol (half-life 3–6 hours), but is consistent with our previous observations of the slow speeds of association and dissociation of timolol with β-adrenoceptors in vitro. There is no evidence, in this study, of β-adrenergic sensitisation following timolol withdrawal, nor of cross-regulation of vascular α1-, α2- or angiotensin II receptors in response to nonselective β-adrenoceptor blockade.