In the EdoxabaN versus warfarin in subjectS UndeRgoing cardiovErsion of Atrial Fibrillation (ENSURE-AF) study, edoxaban was compared with enoxaparin–warfarin in patients who underwent electrical cardioversion of nonvalvular atrial fibrillation, showing comparable low rates of bleeding and thromboembolism. The present study is an ancillary analysis investigating differences in relation to stroke and bleeding risk profiles. It also determined the relation of patients' clinical risk profiles to the quality of anticoagulation control in the warfarin arm. Primary efficacy (composite of stroke, systemic embolic event, myocardial infarction, and cardiovascular death) and safety (composite of major and clinically relevant nonmajor bleeding) outcomes and time to therapeutic range (TtTR) and time in therapeutic range (TiTR) were analyzed in relation to CHA₂DS₂-VASc (congestive heart failure, hypertension, age ≥75 years (2 points), diabetes mellitus, stroke (2 points), vascular disease, age 65–74 years, sex category) and HAS-BLED (hypertension, age, stroke, bleeding tendency/predisposition, labile INRs, elderly age/frailty, drugs such as concomitant aspirin/nonsteroidal anti-inflammatory drugs or alcohol excess) scores. A total of 1,095 patients were randomized to edoxaban and 1,104 received enoxaparin–warfarin. Mean age was 64.3 ± 10 and 64.2 ± 11 years, respectively. Mean CHA₂DS₂-VASc score was 2.6 (standard deviation [SD] 1.5 and 1.4, respectively) and mean HAS-BLED score was 0.9 (SD 0.8) in both arms. There were nonsignificant trends toward lower odds ratios (ORs) for the efficacy end point in patients with CHA₂DS₂-VASc scores >2 and higher ORs with HAS-BLED score ≥3. Mean TiTR was >67%, with no differences between stroke or bleeding risk strata. The correlation between CHA₂DS₂-VASc and TtTR (p = 0.0286) and HAS-BLED and TiTR (p = 0.0286) were statistically significant. In patients who were at high risk of stroke, edoxaban had numerically lower primary efficacy end-point events and showed a trend for higher ORs, with HAS-BLED scores ≥3 compared with enoxaparin–warfarin. TtTR was shorter with higher CHA₂DS₂-VASc scores, whereas bleeding risk was inversely correlated with quality of anticoagulation control.