Intracoronary injection of acetylcholine has been shown to induce coronary spasm in patients with variant angina. To examine its sensitivity and specificity, incremental doses of acetylcholine (20, 50 and 100 micrograms into the left coronary artery and 20 and 50 micrograms into the right coronary artery) were injected into the coronary artery or arteries in 70 patients with variant angina (Group 1) (mean age 57 years) and 93 patients without variant angina or angina at rest (Group 2) (mean age 54 years). Forty patients of the latter group had atypical chest pain, 16 cardiomyopathy, 14 arrhythmia, 11 valvular disease, 7 stable effort angina due to advanced coronary artery disease, 3 congenital heart disease and 2 hypertension. A temporary cardiac pacemaker set at 40 to 50 beats/min was positioned in the right ventricle. Coronary spasm was defined as total occlusion or severe vasoconstriction associated with chest pain or ischemic ST changes on the electrocardiogram or both. In Group 1, acetylcholine induced spasm in 63 (90%) of the 70 patients in the artery or arteries predicted to be responsible for spontaneous attacks. In Group 2, acetylcholine induced coronary spasm only in one patient with effort angina and advanced coronary artery disease although lesser degrees of vasoconstriction (less than or equal to 75% of the luminal diameter) occurred in most patients after acetylcholine (specificity of acetylcholine thus was 99%). In conclusion, intracoronary injection of acetylcholine is sensitive and reliable for the induction of coronary spasm.