Patients with presumed ST-elevation myocardial infarction (STEMI) have no clear culprit artery in approximately 10-15% of cases. We examined the value of cardiac magnetic resonance (CMR) for diagnosis in patients with "no culprit" STEMI. Data from a comprehensive prospective registry of STEMI patients were reviewed from March 2003 to December 2009. "No culprit" patients were followed for diagnosis and clinical outcome. CMR was performed at the discretion of the attending cardiologist. Of 2728 consecutive presumed STEMI patients, 412 (15%) had no clear culprit artery. Of these, 202 (49%) had abnormal cardiac biomarkers with a definitive diagnosis in 157 (78%). Diagnoses in this group included myocardial infarction without a culprit lesion (24%), myopericarditis (22%), and stress cardiomyopathy (21%). In 210 (51%) patients with normal biomarkers, only 84 (40%) received a definitive diagnosis. Diagnoses in this group included myopericarditis (27%), noncardiac causes (21%), and cardiomyopathy (14%). CMR was performed in 123 (30%) "no culprit" patients. Patients who had CMR were more likely to have a definitive diagnosis than those who did not (95/123 [77%] vs. 144/289 [50%]; P=0.01). In particular, "no culprit" patients with abnormal biomarkers were more likely to have a definitive diagnosis with CMR. CMR led to a diagnosis different from the presumptive clinical diagnosis in 53% of all cases. CMR is a valuable diagnostic tool to improve diagnostic accuracy in patients with "no culprit" STEMI.