Seven patients (5 with arteriosclerosis obliterans and 2 with Buerger's disease) completed a two-phase double-blind crossover trial of propranolol in intermittent claudication. Performance was measured on a moving treadmill. In the initial phase, the patients were hospitalized in order to determine an "effective" dose of propranolol. Improvement was noted in all: after 1,600 mg in 5 and after 240 mg and 600 mg in the others. The controlled phase was carried out on an outpatient basis over 8 weeks, the patients receiving propranolol and placebo in a random manner, each for two 2-week periods. Comparison of matched periods of drug and placebo revealed no advantage for propranolol. Patients' performances deteriorated with time. None of the patients evidenced deterioration of occlusive peripheral arterial disease that could be attributed to propranolol, in spite of the high doses used.