Sixty patients (31 male, 29 female) were studied in a 4-week double-blind parallel dose-response study. Patients received amlodipine 1.25 mg (n = 12), 2.5 mg (n = 12), 5 mg (n = 12), 10 mg (n = 12) or placebo (n = 12) once a day. Anti-anginal efficacy was assessed by sequential treadmill testing 24 h post-dose, frequency of anginal attacks and consumption of nitroglycerin, patient and investigator assessment. In the analysis of the final vs baseline total exercise time, the difference between those on amlodipine and those on placebo was significant (P less than 0.01), (all doses). Mean total exercise time increased by 24% in the amlodipine 10 mg group compared with a 20% decrease in the placebo group. The time to onset of angina increased by 37% in the amlodipine 10 mg group, compared with a 16.5% decrease in the placebo group. Differences were statistically significant for the 1.25, 5 and 10 mg amlodipine groups vs placebo. ST-segment charges and rate-pressure product were not affected significantly. There was a statistically significant difference in angina attack frequency and nitroglycerin tablet consumption (P less than 0.05) for all dose groups vs placebo. The majority of patients receiving amlodipine reported improvement in angina symptoms. Side-effects were frequent but mild and dose-related. Amlodipine has significant anti-anginal efficacy. The mechanisms underlying the anti-ischaemic effects of amlodipine are still under discussion.