There are no data concerning the incidence of provoked coronary arterial spasms via intracoronary administration of ergonovine (ER). This study sought to establish the incidence of spasms due to intracoronary injection of ER in Japanese patients who underwent coronary angiography. The subjects were 596 consecutive patients (369 men, mean age 64.2 +/- 10.3 years) who were studied with a selective ER test. ER was administered in total doses of 40 microg into the right coronary artery and 64 microg into the left coronary artery. A positive spasm was defined as a total or subtotal occlusion. Coronary vasospasms were determined in 173 patients (29.0%). Spasms occurred often in patients with ischemic heart disease (43.3%); during effort and rest in patients with angina (46.3%), exertional angina (27.7%), recent myocardial infarction (36.7%), healed myocardial infarction (34.1%), and especially in patients with rest angina (55.5%), but were relatively uncommon in patients with nonischemic heart disease (3.7%). The incidence of provoked coronary spasms in this study was 2.2-2.6 times higher than in previous reports with intravenous ER administration. More spasms were superimposed on significant atherosclerotic lesions than on nonfixed atherosclerotic lesions (42.8% vs 24.0%, p < 0.01). No serious or irreversible complications were observed in this study. In conclusion, intracoronary administration of ER was a safe and reliable test. Compared with Caucasian patients, in Japanese patients, coronary arterial spasms occurred 2-3 times more frequently with various cardiac disorders.