Abstract Funding Acknowledgements Type of funding sources: None. Background Cardiac masses are an heterogeneous group of very rare formations, representing a major diagnostic challenge given the difficulty often encountered in assessing their nature through imaging examinations. Cardiac Magnetic Resonance (CMR) is a noninvasive key diagnostic tool providing anatomical, functional, and tissue characteristic information. However data validating this imaging approach are limited. Purpose 1) to analyze the clinical and CMR features of patients with benign and malignant tumors and thrombus; 2) to asses the diagnostic accuracy of CMR in comparison with the histologic examination;3) to evaluate the prognostic rule of clinical and CMR features in predicting the primary endpoint of all cause mortality. Methods 92 Patients undergoing CMR for suspected cardiac mass between June 2004 and January 2022 were retrospectively evaluated. Patients with no mass 8 (9%) or pseudomass 11 (12%) were excluded. Based on CMR images, masses were divided into benign tumor, malignant tumor, and thrombus. Clinical, CMR and histological data were collected. Results 73 patients (mean age: 62±15; 51% female) with masses were finally enrolled, 27 (37%) with a diagnosis at CMR of benign tumor, 22 (30%) with malignant tumor, and 24 (33%) with thrombus. Several clinical and CMR variables were significantly associated with one type of mass rather than another. Among clinical variables the history of malignancy and smoking were seen in patients with malignant tumors (p<0,002) while dyspnea and chest pain were significantly found in patients with thrombi (p<0,016). Among CMR features high size, infiltration, pericardial repeats, pericardial effusion, signal inhomogeneity, First Pass Perfusion (FPP) and Late Gadolinium Enhancement (LGE) were associated with malignant masses (p = 0.001), while reduced pre-contrast visualization, altered kinetics, low ejection fraction (EF) and long-TI were observed in thrombi (p = 0.001). CMR was accurate in 91.2% (31/34) of patients in comparison with the histological examination. Accuracy for benign tumors was 91.3% (AUC=0.913), for malignant tumors 94% (AUC=0.940) and for thrombi 83% (AUC=0.833). Patients diagnosed at CMR with thrombus had similar mortality as those with benign tumor (p = 0,678) while patients with malignant lesions had higher mortality (HR: 4.98 [95% CI: 1.11–22.2], p = 0,035). Regardless of diagnosis at CMR, age, tumor history, mass size, signal inhomogeneity, presence of FPP, pericardial effusion, and pericardial repeats were found to predict long-term mortality (p<0.017). Conclusions Although CMR has high diagnostic accuracy, histologic examination remains the diagnostic gold standard in determining the type of cardiac mass. In addition to playing a key role in the diagnostic process, CMR is also useful in predicting the clinical outcome of patients with cardiac masses.