In a randomized double-blind trial performed in 58 offices of internists and general practitioners, Persumbran (25 mg dipyridamole + 10 mg oxazepam) was compared with placebo. A total of 344 patients with mild-to-moderate angina pectoris in whom the tranquilizing effect of oxazepam appeared meaningful and who had already previously been treated with Persumbran, were admitted to the study. In more than 50% of the patients, coronary heart disease had been confirmed by ergometry or angiography. The dose administered during the trial was oriented to the previous treatment with Persumbran. During the first week of the trial, all patients received a placebo; in the following 6 weeks, one group of patients received Persumbran, the other was continued on placebo. If required, nitroglycerine capsules were to be allowed. The patients in the Persumbran group registered an average of 18% fewer "mild", 50% fewer "moderate", and 35% fewer "severe" attacks of angina pectoris; the total number of attacks, weighted for severity, was significantly lower than in the placebo group. The use of nitroglycerine capsules was 39% less in the Persumbran group as compared with the placebo group. Also episodes of "prolonged cardiac complaints whose duration was not exactly definable" were significantly less frequent in the Persumbran group. Moreover, restrictions in activity due to cardiac complaints, as recorded by the patients were significantly fewer in the Persumbran group.