We examined the role of prosthesis–patient mismatch on left ventricular mass regression after aortic valve replacement for chronic aortic valve regurgitation.We selected patients who had complete preoperative and follow-up echocardiograms with measurement of left ventricular mass. Patients were excluded who had moderate or greater aortic valve stenosis, concomitant coronary artery bypass grafting, or mitral valve procedures.Patients’ mean age was 55 ± 17 years; 21% were female. The mean preoperative indexed left ventricular mass was 150 ± 45 g/m 2 . Patients with mildly (n = 44; mean indexed mass, 126 ± 15 g/m 2 ), moderately (n = 31; mean indexed mass, 168 ± 11 g/m 2 ), or severely (n = 15; mean indexed mass, 241 ± 34 g/m 2 ) increased preoperative indexed left ventricular mass, were similar, except for lower ejection fractions, larger end-diastolic dimensions, and larger ventricular wall thicknesses in the severely enlarged group (P < .001). Thirteen patients had prosthesis–patient mismatch and were similar to patients without prosthesis–patient mismatch, except for a greater body surface area, fewer mechanical valves, and smaller valve sizes in those with prosthesis–patient mismatch (P < .05). At a mean follow-up of 3.2 ± 2.4 years, the average reduction in indexed left ventricular mass was 50 ± 38 g/m 2 ; late mass regression was unrelated to labeled valve size, prosthesis–patient mismatch, or measured indexed effective aortic valve area. A greater preoperative indexed left ventricular mass (P < .001) was an independent predictor of greater left ventricular mass regression. Despite having greater left ventricular mass regression, patients with severe preoperative indexed left ventricular mass did not return to normal values (mean, 142 ± 25 g/m 2 ).Left ventricular mass regression after aortic valve replacement for chronic aortic regurgitation is unrelated to indexed prosthetic valve area. Although incomplete, regression is greatest in patients with the largest preoperative indexed left ventricular mass.