We reviewed our experience with the ergonovine provocative test for coronary arterial spasm in 40 patients with pain in the chest believed to be angina pectoris and in one patient with a myocardial infarction and normal coronary arteries. Twenty-nine patients had normal coronary arteries, while 12 had mild to moderate lesions. Ergonovine maleate was administered incrementally in total cumulative doses of 0.25 mg to 1.2 mg. The effect of ergonovine on coronary arterial caliber was determined by comparing the arterial diameter from the angiogram obtained after administration of ergonovine with that from the control. Measurements were made at the same preselected points in both films and also at points of greatest response. Excluding the three cases with complete occlusion, the mean reduction in coronary arterial diameter at preselected points was 12 +/- 15 percent. When the points of greatest response were examined, the maximum reduction in coronary arterial diameter was less than 25 percent in 13 patients, 25 to 50 percent in 20 patients, and more than 50 percent in eight patients. The patterns of response included complete occlusion of a vessel in the three patients with variant angina, diffuse narrowing in 16, diffuse and focal narrowing in six, and spasm at the catheter tip in three patients. All patients with maximum reductions of more than 50 percent in coronary arterial diameter and six of those with maximum reductions of 25 to 50 percent had pain in the chest, but only the three with complete occlusion had associated changes in the S-T segment. Thus, the response in patients with variant angina represents one end of a spectrum of responses to administration of ergonovine. In addition, a large number of patients may have ergonovine-induced pain in the chest without electrocardiographic changes and only an intermediate degree of coronary arterial spasm.