Background: EGPA is a rare small-vessel vasculitis. Cardiovascular (CV) involvement is an important predictor of adverse outcomes but is challenging to detect. CMR is as an excellent tool to detect cardiac EGPA but is not widely available. It is unknown who would benefit most from comprehensive CV assessment with CMR. We aimed to determine the prevalence of cardiac EGPA with multiparametric CMR and to identify potential clinical/biochemical predictors of CV involvement. Methods: This single-center study was performed in EGPA patients who underwent CMR as part of routine clinical evaluation including comprehensive tissue characterization with T1 mapping, T2-STIR, T2 mapping, and late gadolinium enhancement imaging, in accordance with SCMR guidelines. Results: Sixty-three EGPA patients were included (Table 1). CV involvement (defined by typical LGE, LV dysfunction or pericardial inflammation) was seen in 37%. The median LV and RV ejection fraction was 58% and 57%. Two (3%) patients demonstrated increased T2 signal and mapping values, suggesting acute cardiac injury. LGE was present in 18 (29%) patients with a predominant pattern of subendocardial fibrosis. Patients with CV involvement were younger when diagnosed with EGPA, had higher NT-pro-BNP level (sensitivity 30%, specificity 98%) and lower EF (sensitivity 48%, specificity 100%) (Table 1). Troponin levels were not predictive of CV involvement (Table 1) as defined by positive CMR. Conclusion: Clinical and biochemical markers had low sensitivity for detecting cardiac involvement which was present in over a third of patients. CMR is necessary to detect cardiac involvement and should be considered part of the baseline evaluation of patients with EGPA.