Abstract Background Previously we reported that incident rate of heart failure (HF) hospitalization was higher in female than male and stroke was comparable between female and male in Japanese atrial fibrillation (AF) patients. We also reported cardiovascular death was mainly related to HF but not stroke. However, the gender difference in cause of death in Japanese AF patients is little known. Method The Fushimi AF Registry is a community-based prospective survey of AF patients in Fushimi-ku, which is a typical urban district of Japan. We started to enroll patients from March 2011, and follow-up data were available for 4,496 patients by the end of February 2022. In the entire cohort, 2,677 were male and 1,819 were female. We compared clinical characteristics and cause and incidence rate of death between two groups. Result Female patients were older (female vs. male; 76.7 years vs. 71.5 years, p<0.01), less in body weight (51.2 kg vs. 65.0 kg, p<0.01), had higher pulse rate (79.4 bpm vs. 77.5 bpm, p<0.01), less often persistent/permanent type (48.2% vs. 52.0%, p=0.04), less likely to have previous stroke/systemic embolism (18.6% vs. 21.1%, p=0.04), diabetes mellitus (19.9% vs. 26.2%, p<0.01), coronary artery disease (12.3% vs. 16.0%, p<0.01) and more likely to have prior HF (32.2% vs. 24.0%, p<0.01), chronic kidney disease (39.3% vs. 33.7%, p<0.01)and had higher CHADS2 score (2.15 vs. 1.96, p<0.01). In prescription data, female patients less often received oral anticoagulant (52.1% vs. 58.2%, p<0.01) and antiplatelet drug (23.8% vs. 29.0%, p<0.01). Systolic blood pressure, hypertension and major bleeding were comparable between two groups. During the median follow-up of 2,093 days, the incidence rate of all cause death was comparable between female group and male group (5.20 vs. 4.86 % per person-year, log rank p=0.32). The incidence rate of cardiac death was significantly higher in female group than male group (0.98 vs. 0.70 % per person-year, log rank p=0.02) and the incidence rate of vascular death was also significantly higher in female group than male group (0.43 vs. 0.26 % per person-year, log rank p=0.03). The incidence rate of death due to HF, which was main cause of cardiac death, was higher in female group but not significant (0.86 vs. 0.66 % per person-year, log rank p=0.08). We analyzed the proportion of cause of death in each gender, cardiac death and stroke were higher and non-cardiovascular death was lower in female than male group. Conclusion The incidence rates of cardiac death and vascular death were significantly higher in female group than male group, and death due to HF was non-significantly higher in female group. Concerning the proportion of cause of death in each gender, cardiac death and stroke were higher in female than male. The impact of cardiovascular events may be stronger and prevention of them may be more important in female.