There is ambiguity whether frail patients with atrial fibrillation (AF) managed with vitamin K antagonists (VKAs) should be switched to a non-vitamin K oral anticoagulant (NOAC). We conducted a pragmatic, multicenter, open-label, randomized controlled superiority trial. Older AF patients living with frailty (age ≥75 years plus a Groningen Frailty Indicator (GFI) score ≥3) were randomized to switch from INR-guided VKA treatment to a NOAC or to continued VKA treatment. Patients with a glomerular filtration rate <30 mL/min/1.73 m or with valvular AF were excluded. Follow-up was 12 months. The cause-specific hazard ratio (HR) was calculated for occurrence of the primary outcome which was a major or clinically relevant non-major bleeding complication, whichever came first, accounting for death as a competing risk. Analyses followed the intention-to-treat principle. Secondary outcomes included thromboembolic events. Between January 2018 and June 2022, a total of 2,621 patients were screened for eligibility and 1,330 patients were randomized (mean age 83 years, median GFI 4). After randomization 6 patients in the switch to NOAC arm and 1 patient in the continue with VKA arm were excluded due to the presence of exclusion criteria, leaving 662 patients switched from a VKA to a NOAC and 661 patients continued VKAs in the intention-to-treat population. After 163 primary outcome events (101 in the switch arm, 62 in the continue arm), the trial was stopped for futility according to a prespecified futility analysis. The HR for our primary outcome was 1.69 (95% CI 1.23-2.32). The HR for thromboembolic events was 1.26 (95% CI 0.60 to 2.61). Switching INR-guided VKA treatment to a NOAC in frail older patients with AF was associated with more bleeding complications compared to continuing VKA treatment, without an associated reduction in thromboembolic complications.