The pathogenesis of dilated cardiomyopathy (DCM) is unknown, but clinical evidence suggests that coronary vasospasm is associated with the development of DCM in some cases. In the present study, we aimed to clarify the prevalence of coronary vasospasm in patients with DCM, the characteristics of patients with DCM and coronary vasospasm, and the effects of anti-vasospastic agents on patients with DCM and coronary vasospasm. This study included 18 consecutive patients with DCM who underwent cardiac catheterization with ergonovine provocation testing. The patient was diagnosed as having coronary vasospasm if ergonovine induced coronary vasoconstriction > or = 75% diameter narrowing was observed compared to the diameter after nitroglycerin administration. Six (33%) patients were found to have coronary vasospasm and anti-vasospastic agents were added after the cardiac catheterization. The prevalence of atrial fibrillation in the patients with DCM and coronary vasospasm was greater than that in DCM without coronary vasospasm [67% vs 8% (P<0.05)]. The left ventricular end-diastolic dimension decreased from 61 mm (56/64) to 55 mm (53/56) (median, 25th/75th percentile, P<0.05) and the left ventricular ejection fraction increased from 36% (32/40) to 47% (46/48) (median, 25th/ 75th percentile, P<0.05) after the administration of anti-vasospastic agents and 4 of the 6 patients improved symptomatically. Therefore, ergonovine provocation testing is useful in identifying patients with DCM and coronary vasospasm, in whom cardiac performance is expected to be improved with anti-vasospastic agent therapy. DCM patients with atrial fibrillation may be a clue for identifying patients with coronary vasospasm.