Abstract Background  Western guidelines recommend an international normalized ratio (INR) range of 2 to 3 when using warfarin for stroke prevention in atrial fibrillation (AF), but lower INR ranges are frequently used in East Asia. We performed a systematic review and meta-analysis of randomized controlled trials (RCTs) in AF patients comparing the effect of lower versus standard INR targets on thromboembolism, major bleeding, and mortality. Methods  We searched Western databases including Cochrane CENTRAL, Medline, and Embase as well as Chinese databases including SinoMed, CNKI, and Wanfang Data. We pooled risk ratios (RRs) using random-effects model. We grouped INR targets in two ways: (1) any study-specific lower versus standard targets and (2) INR ranges of approximately 1.5 to 2 versus 2 to 3. Results  Seventy-nine RCTs ( n  = 12,928) met eligibility criteria: 74 ( n  = 11,322) from East Asia and 5 ( n  = 1,606) from Western countries. Compared with standard targets, lower INR ranges were associated with higher rates of thromboembolism (76 RCTs, n  = 12,577: 7.1% vs. 4.4%, RR 1.50, 95% confidence interval CI 1.29–1.74, I 2  = 0%), lower rates of major bleeding (61 RCTs, n  = 10,815: 2.2% vs. 4.4%, RR 0.54, 95% CI 0.44–0.67, I 2  = 0%), and similar mortality (32 RCTs, n  = 7,327: 4.8% vs. 5.2%, RR 1.00, 95% CI 0.85–1.19, I 2  = 0%). Results were similar when comparing target ranges of approximately 1.5 to 2 versus 2 to 3. Conclusion  Moderate quality evidence suggests lower INR targets reduce bleeding but increase thromboembolism in AF. The data are dominated by East-Asian studies, limiting generalizability to Western populations. Until higher quality data demonstrate otherwise, an INR range of 2 to 3 should remain standard for thromboembolic prophylaxis in AF.