Cardiac magnetic resonance (CMR) offers the capability to objectively detect pericarditis by identifying pericardial thickening, edema/inflammation by Short-TI Inversion Recovery-T2 weighted (STIR-T2w) imaging, edema/inflammation or fibrosis by late gadolinium enhancement (LGE), and presence of pericardial effusion. This is especially helpful for the diagnosis of recurrent pericarditis. Aim of the present paper is to assess the diagnostic accuracy of CMR findings as well as their potential prognostic value for the diagnosis of recurrent pericarditis. Multicenter cohort study of consecutive patients with recurrent pericarditis evaluated by CMR. We included 128 consecutive cases (60 males, 47%; mean age 48 ± 14 years). CMR was performed at a mean time of 12 days (95% confidence interval 15 to 21) after the clinical diagnosis. We evaluated the diagnostic accuracy and areas under the receiver operating characteristic (ROC) curve for CMR diagnostic criteria and complications (additional recurrences, cardiac tamponade, and constrictive pericarditis). Areas under the ROC curve were respectively 64% for pericardial thickening, 84% for pericardial edema, 82% for pericardial LGE, and 71% for pericardial effusion. After a mean follow-up of 34 months, recurrences occurred in 52% of patients, tamponade in 6%, and constrictive pericarditis in 11%. Using a multivariable Cox model, elevation of CRP and presence of CMR pericardial thickening were predictors of adverse events, whereas the presence of CMR LGE was associated with a lower risk. The prognostic model for adverse events using gender, age, CRP level, and all CMR variables showed a C-index of 0.84. In conclusion, CMR findings show high diagnostic accuracy and may help identifying patients at higher risk of complications.