Abstract Background Atrial fibrillation (AF) is common in older people with frailty and is associated with an increased risk of stroke and systemic embolism. Whilst oral anticoagulation is associated with a reduction in this risk, there is a lack of data on the safety and efficacy of direct oral anticoagulants (DOACs) in people with frailty. This study aims to report clinical outcomes of patients with AF in the Effective Anticoagulation with Factor Xa Next Generation in Atrial Fibrillation–Thrombolysis in Myocardial Infarction 48 (ENGAGE AF-TIMI 48) trial by frailty status. Methods Post hoc analysis of 20,867 participants in the ENGAGE AF-TIMI 48 trial, representing 98.8% of those randomised. This double-blinded double-dummy trial compared two once-daily regimens of edoxaban (a DOAC) with warfarin. Participants were categorised as fit, living with pre-frailty, mild-moderate, or severe frailty according to a standardised index, based upon the cumulative deficit model. The primary efficacy endpoint was stroke or systemic embolism and the safety endpoint was major bleeding. Results A fifth (19.6%) of the study population had frailty (fit: n = 4459, pre-frailty: n = 12,326, mild-moderate frailty: n = 3722, severe frailty: n = 360). On average over the follow-up period, the risk of stroke or systemic embolism increased by 37% (adjusted HR 1.37, 95% CI 1.19–1.58) and major bleeding by 42% (adjusted HR 1.42, 1.27–1.59) for each 0.1 increase in the frailty index (four additional health deficits). Edoxaban was associated with similar efficacy to warfarin in every frailty category, and a lower risk of bleeding than warfarin in all but those living with severe frailty. Conclusions Edoxaban was similarly efficacious to warfarin across the frailty spectrum and was associated with lower rates of bleeding except in those with severe frailty. Overall, with increasing frailty, there was an increase in stroke and bleeding risk. There is a need for high-quality, frailty-specific population randomised control trials to guide therapy in this vulnerable population. Trial registration ClinicalTrials.gov NCT00781391. First registered on 28 October 2008