Objective: To evaluate warfarin use in Chinese patients with atrial fibrillation (AF) undergoing percutaneous coronary intervention (PCI) by investigating the stroke and major adverse cardiac and cerebral events (MACCEs) and bleeding events. Methods: Retrospective cohort study of the 5 year follow-up of 1134 patients with AF who underwent PCI. The patients were grouped according to whether they received warfarin or not. Baseline characteristics and the occurrence of MACCEs and bleeding events were compared between the two groups using the CHA2DS2-VASc and HAS-BLED scoring. Cox regression analysis was used to identify factors related to the occurrence of MACCEs and bleeding. Results: Overall MACCE (p = .008) and mortality (p = .004) rates were significantly lower in the warfarin group compared with the non-warfarin group. Major bleeding, minor bleeding and overall bleeding were comparable in the two groups. Recurrent myocardial infarction (HR = 10.129, 95% CI = 4.737-21.655; p < .001) and a baseline CHA2DS2-VASc score >4 (HR = 2.035, 95% CI = 1.121-3.692; p = .019) were independent predictors of MACCEs in the warfarin group. A baseline HAS-BLED score >= 3 (HR = 5.498, 95% CI = 3.773-8.013; p < .001) and previous bleeding (HR = 3.058, 95% CI = 1.319-7.088; p = .009) were independent predictors of bleeding. Conclusions: Warfarin reduces the incidence of MACCEs but does not increase bleeding events in Chinese patients with AF who underwent PCI. For patients taking warfarin, recurrent myocardial infarction and a baseline CHA2DS2-VASc score >4 were related to MACCE occurrence.