Abstract.To evaluate the effect of metoprolol on insulin sensitivity and diurnal plasma hormone levels, seven mildly hypertensive subjects were investigated (four men and three women, age 52 Ϯ 8, body mass index 25 . 4 Ϯ 1 . 9, mean Ϯ SD). The study had a placebo-controlled, double-blind, crossover design with 6 weeks' metoprolol treatment (100 mg b.i.d) vs. placebo. At the end of each treatment period 24-h blood samples were collected continuously for diurnal analysis of hormone levels and a hyperinsulinaemic euglycaemic clamp combined with [3-3 H]-D-glucose infusion was performed. Insulin sensitivity was evaluated by means of three different methods: diurnal plasma insulin and glucose levels; glucose consumption; and insulin sensitivity index during euglycaemic clamp conditions. Fasting blood glucose and insulin concentrations as well as mean plasma diurnal levels of insulin, growth hormone, testosterone and cortisol were similar after placebo and metoprolol treatment, whereas noradrenaline and adrenaline levels were significantly increased after metoprolol. During the clamp, plasma insulin was significantly higher after metoprolol treatment than after placebo treatment (56 Ϯ 3 vs. 64 Ϯ 2 mU L -1 , P < 0 . 05). Consequently, the insulin sensitivity index [glucose infusion rate (GIR)/ plasma insulin] was lower after metoprolol treatment (16 . 1 Ϯ 2 . 6 vs. 10 . 2 Ϯ 1 . 2, P < 0 . 05), although GIR was not significantly changed. We suggest that the insulin sensitivity index may not accurately reflect the insulin effect as the plasma level of insulin was significantly increased during insulin infusion but not at 24 h, possibly because of alteration of distribution and/or degradation rate of exogenous insulin. Thus, the likelihood of metoprolol inducing insulin resistance in hypertensive subjects may be less than previously proposed.