This paper assesses mechanisms that may contribute to the higher incidence of increased muscle fatigue during exercise and reduced exercise performance as observed with selective compared with non-selective beta-adrenoceptor antagonists. Published data and the results obtained in 8 healthy subjects (mean age 23 years) studied before and after acute beta-adrenoceptor blockade with pindolol (nonselective, 10 mg) and metoprolol (beta 1-receptor selective, 100 mg) suggest that the differences in the cardiovascular and respiratory effects between the 2 types of antagonists are marginal and cannot explain the discrepancies concerning exercise perception and performance. Conversely, basic differences between the 2 types of antagonists were shown in different groups of hypertensive men (mean age 32 years) studied before and after 4 weeks of treatment with pindolol (15 mg), and with metoprolol (200 mg) and acebutolol (cardioselective, 500 mg), by single crossover technique. Whereas lipolysis was similarly inhibited by both selective and non-selective antagonists, hypoglycaemia occurred only under non-selective blockade. It apparently reflects the inhibition of glycogen breakdown; concomitant rises in plasma adrenaline and ACTH probably reflect counter-regulatory mechanisms.