The use of beta-blockers in diabetes mellitus has largely been restricted because of the reported adverse effects. Clinical investigations aimed at elucidating the possible reactions associated with the use of beta-blockers have disclosed no evidence of masking or signs or insulin-induced hypoglycaemia or potentiation of the insulin effect. Prolonged hypoglycaemia may develop, however, as a result of physical effort. There is no proof that during insulin-induced hypoglycaemia the concentrations of counter-regulatory hormones are depressed, but that of glycerol, a gluconeogenic precursor, is slightly diminished. Intensification of the hypertensive reaction during hypoglycaemia is less likely to occur during treatment with beta-selective blockers. In insulin-dependent diabetics receiving beta 1-blockers there is no evidence of any change - either deterioration or improvement - in metabolic control. In one small controlled trial there was no sign of impairment of the peripheral arterial circulation over a short period of administration of a non-selective beta-blocker. In general, for patients suffering from insulin-dependent diabetes, cardioselective agents are preferable. Since cardioselectivity is a dose-dependent property, reasonable caution should also be observed when using this type of drug in diabetes.