AIMS: Haemodynamics alone do not fully explain symptoms and prognosis in clinically severe aortic stenosis (AS). Myocardial disease, specifically diffuse myocardial fibrosis (DMF), may contribute. We used equilibrium contrast cardiovascular magnetic resonance (EQ-CMR) and sought to non-invasively measure DMF in severe AS and determine its clinical significance before and after valve replacement. METHODS AND RESULTS: Patients with severe AS underwent echocardiography, brain natriuretic peptide (BNP), 6 min walk test (6MWT), and EQ-CMR pre- (n = 63) at baseline and at 6 months post- (n = 42) aortic valve replacement (AVR). EQ-CMR was also performed in 30 normal controls. Baseline: patients with AS had more DMF than controls (18 vs. 13%, P = 0.007) with a wide range (5-38%) that overlapped controls. The extent of diffuse fibrosis correlated inversely with the 6MWT performance (r(2) = 0.22, P = 0.001). Those with severe diastolic dysfunction had more DMF (P = 0.01). On multivariable analysis, the predictors of performance at 6MWT were diffuse fibrosis and BNP (P = 0.003 and 0.02, respectively). Post-op: following valve replacement, morphological and functional parameters improved [6 MWT, LA area, BNP, left ventricular (LV) hypertrophy, and volumes]. LV hypertrophy regression was shown to be cell volume reduction (P < 0.001) and not fibrosis regression (P = 0.54). Of the five deaths over six-month follow-up, four occurred in patients in the highest tertile of DMF. CONCLUSION: DMF as measured by EQ-CMR is elevated in severe AS vs. normal controls but with a considerable overlap. It correlates with functional capacity at baseline. LV hypertrophy regression 6 months after AVR is cellular rather than fibrosis resolution.