Left atrial appendage (LAA) obliteration is a proven stroke-preventive measure for patients with nonvalvular atrial fibrillation (AF). However, the efficacy of LAA obliteration for patients with AF after bioprosthetic mitral valve replacement (MVR) remains unclear. This study aimed to estimate the efficacy of LAA obliteration in preventing embolism and to investigate the predictors of thromboembolism after bioprosthetic MVR. We retrospectively studied 173 AF subjects with bioprosthetic MVR; among them, 81 subjects underwent LAA obliteration using an endocardial running suture method. The main outcome measure was the occurrence of thrombosis events (TEs). The mean follow-up time was 40 +/- 17 months. AF rhythm was observed in 136 patients postoperatively. The incidence rate of TEs was 13.97 % for postoperative AF subjects; a dilated left atrium (LA; > 49.5 mm) was identified as an independent risk factor of TEs (OR = 10.619, 95 % CI = 2.754-40.94, p = 0.001). For postoperative AF patients with or without LAA, the incidence rate of TEs was 15.8 % (9/57) and 12.7 % (10/79; p = 0.603), respectively. The incidence rate of TEs was 2.7 % (1/36) and 4.2 % (2/48) for the subgroup patients with a left atrial diameter of < 49.5 mm, and 38.1 % (8/21) and 25.8 % (8/31) for those with a left atrial diameter of > 49.5 mm (p = 0.346). Surgical LAA obliteration in patients with valvular AF undergoing bioprosthetic MVR did not reduce TEs, even when the CHA(2)DS(2)-VASc score (a score for estimating the risk of stroke in AF) was a parts per thousand yenaEuro parts per thousand 2 points.