A cohort of thirty-five patients with angina pectoris were followed-up for six months with placebo as the only anti-anginal treatment apart from short-acting nitroglycerin. Only patients with at least five attacks per week were entered into the study. Prior to the study, their average number of attacks per week was 10.3(+/- 0.9) and their average nitroglycerin tablet consumption per week was 10.6(+/- 1.2). After admission into the study, the individual dosage of placebo could be blindly titrated during the first eight weeks. Once an adequate dosage was found, it was continued for 6 months. The primary end-point was predefined success or failure of the placebo treatment. Failure was assessed on the lack of improvement, occurrence of 'side-effects', or need for other anti-anginal treatment. In 27 patients, the placebo treatment was said to be a success (95% confidence limits of the percentage of success: 60-90%). No severe cardiac event occurred (confidence limits: 0-10%). Six patients developed 'side-effects'. The number of attacks per week decreased by 48% (P less than 0.0001) during the titration period (8 weeks) and by 77% during the whole 6 months. Exercise test improved, however, less markedly.