Objective: To define the diagnostic yield of cardiac magnetic resonance (CMR) in differentiating the underlying causes of myocardial infarction with nonobstructive coronary arteries (MINOCA) and to determine the long-term prognostic implications of such diagnoses. Methods: Cardiac magnetic resonance evaluation was performed in 227 patients (mean age, 56.4±14.9 years; 120 [53%] female) with a “working diagnosis” of MINOCA as defined by presentation with a troponin-positive acute coronary syndrome (troponin I >0.04 μg/L) and nonobstructed coronary arteries between January 1, 2007, and February 28, 2013. Follow-up was performed to assess the primary composite end point of myocardial infarction, heart failure, and all-cause mortality. Results: Cardiac magnetic resonance identified nonstructural cardiomyopathies in 97 (43%) patients, myocardial infarction in 55 (24%) patients, structural cardiomyopathies in 27 (12%) patients, and pulmonary embolism in 1 patient. No CMR abnormalities were identified in the remaining patients. Kaplan-Meier analysis demonstrated the ability of a CMR diagnosis to predict the risk of the primary composite end point (P=.005) at 5-year follow-up. Worse outcomes were seen among patients with “true” MINOCA and a normal CMR image compared with those with CMR-confirmed myocardial infarction (P=.02). Use of antiplatelets (78% [37/45] vs 95% [52/55]; P=.01), beta blockers (56% [25/45] vs 82% [45/55]; P=.004), and statins (64% [29/45] vs 85% [47/55]; P=.01) was significantly lower in patients with true MINOCA with normal CMR imaging compared with those with CMR-confirmed myocardial infarction. Conclusions: Cardiac magnetic resonance carries a high diagnostic yield in patients with MINOCA and predicts long-term prognosis. Patients with MINOCA with normal CMR imaging had an increased rate of major adverse cardiac events and lower use of guideline-recommended myocardial infarction therapy compared with those with CMR-confirmed myocardial infarction.