This study aimed to evaluate the safety and efficacy of vitamin K antagonist (VKA) in atrial fibrillation (AF) patients with previous ulcer bleeding.In this multicenter, retrospective analysis, clinical outcomes of 754 AF patients with a history of ulcer bleeding were evaluated. After ulcer treatment, 458 patients (61%) were treated with VKA, and the outcomes were compared to 296 patients (39%) without VKA.VKA treatment significantly increased major bleeding (7.3%/year vs 3.2%/year, P < 0.001), and reduced major adverse cardiac events (MACE) (5.4%/year vs 10.0%/year, P < 0.001). Specifically, risk of gastrointestinal bleeding was significantly higher in the VKA group than no-VKA group (5.7%/year vs 2.6%/year, P < 0.001). Consequently, there was no difference in the incidence of composite of a MACE and major bleeding, between the 2 groups. In patients with time in the therapeutic range (TTR) ≥65%, VKA significantly decreased MACE (2.8%/year vs 10.0%/year, P < 0.001) without increasing major bleeding. Net clinical benefit model showed beneficial effects of VKA in patients with TTR ≥65%, and harmful effects in those with TTR < 55%.In AF patients with previous ulcer bleeding, VKA treatment did not improve clinical outcomes unless the international normalized ratio level was constantly maintained (TTR ≥65%), as the gastrointestinal bleeding (GIB) risk significantly increased.