In order to evaluate the effects of acupuncture in moderate, stable angina pectoris, 49 patients were randomized to either genuine or sham acupuncture. In sham acupuncture needles were inserted into points within the same spinal segment as in genuine acupuncture, but outside the Chinese meridian system. The effect was evaluated from exercise tests, anginal attack rate and nitroglycerin consumption. There were no significant differences between the effects of genuine and sham acupuncture either on exercise test variables or on subjective variables. In patients receiving genuine acupuncture there was a significant increase in exercise tolerance (median 9%) and in delay of onset to pain (median 10%). No significant changes were observed in patients receiving sham acupuncture. Within both groups there was a median reduction of 50% in anginal attack rate and nitroglycerin consumption, and there was no significant difference between the results achieved in the two groups. It is concluded that with the present design it was not possible to demonstrate any significant differences between the effect of genuine and sham acupuncture.