OBJECTIVE: To investigate the characteristics and outcome of abnormal vaginal bleeding in women receiving edoxaban or warfarin for treatment of venous thromboembolism (VTE). DESIGN AND SETTING: Post hoc analysis of the Hokusai-VTE study, a multicentre, randomised, double-blind trial comparing edoxaban with warfarin for acute symptomatic VTE. POPULATION: Women below 50 years receiving edoxaban or warfarin for treatment of VTE. METHODS: We collected data on diagnostic measures, treatment, and clinical outcome of abnormal vaginal bleeding events. MAIN OUTCOME MEASURES: Occurrence of major and clinically relevant nonmajor (CRNM) abnormal vaginal bleeding events. RESULTS: In all, 628 women aged under 50 years were treated with edoxaban and 665 with warfarin. The rate of abnormal vaginal bleeding was 15/100 person-years (py) (95% CI 11-19) in women receiving edoxaban and 9/100 py (95% CI 6-12) in the warfarin group (hazard ratio: 1.7, 95% CI 1.1-2.5). Major abnormal vaginal bleeding occurred in eight (1.3%) women on edoxaban and in three (0.9%) women receiving warfarin [odds ratio (OR) 2.8; 95% CI 0.8-10.8], and CRNM abnormal vaginal bleeding occurred in 53 (8.4%) women treated with edoxaban and in 37 (5.6%) on warfarin therapy (OR 1.6, 95% CI 1.0-2.4). Over 85% of all vaginal bleeds were characterised by heavy menstrual bleeding. Major bleeds frequently required treatment, and in more than 75% of patients anticoagulant therapy was adjusted. The severity of clinical presentation and course of major and CRNM bleeds was mild in most patients. CONCLUSIONS: Abnormal vaginal bleeding occurred more frequently in women treated with edoxaban than with warfarin. Reassuringly, most events could be managed conservatively and had a mild outcome. TWEETABLE ABSTRACT: Abnormal vaginal bleeding occurred more frequently in women treated with edoxaban than with warfarin.