Simple tests of verbal reasoning and mental arithmetic, taken under mildly stressful conditions, have been shown to give a reproducible test of intellectual function within groups of normal subjects. Using these tests, in two separate examinations, a double-blind cross-over study was performed on 35 medical students to assess the effects of acute beta-adrenoceptor blockade with propranolol on intellectual function. With placebo treatment, students recorded an average total score of 231.3 marks, with average scores of 108.9 marks on the mental arithmetic paper and 122.4 marks on the verbal reasoning paper. Treatment with propranolol was associated with an improvement in total score of 9.2 +/- 3.9 marks (P less than 0.05), an improvement in mental arithmetic score of 5.6 +/- 2.3 marks (P less than 0.05) and an improvement in verbal reasoning score of 3.6 +/- 2.4 marks (NS). Eighteen out of the 35 students said that they were mildly anxious before one examination and 13 students said they were anxious before both examinations. Those students who admitted anxiety seemed to benefit the most, in terms of improved examination performance, from treatment with propranolol.