Although many patients with long-term oral anticoagulation (OAC) can manage their medication safely and reliably themselves, no study on elderly patients has as yet assessed the safety and efficacy of OAC self-management with major thromboembolic and haemorrhagic complications as primary outcomes. In this multi-centre trial, patients aged 60 years or more were randomised into a self-management (SMG) (N = 99) or routine care group (RCG) (N = 96). The primary outcome was the combined endpoint of all thromboembolic events requiring hospitalisation and all major bleeding complications. Mean follow-up was 2.9 +/- 1.2 and 3.0 +/- 1.1 years in the SMG and RCG, respectively. In intention-to-treat analysis, 12 patients in the SMG versus 22 patients in the RCG reached a primary endpoint (hazard ratio [HR]: 0.50; 95% confidence interval [CI]: 0.25 to 1.00; p = 0.049). The post-hoc analysis on OAC treatment sensitivity supported the benefit of self-management (HR 0.27; 95% CI: 0.13 to 0.71; p = 0.006). Fifteen patients in the SMG died during the study, but none of the deaths was directly associated with anticoagulation therapy. In the RCG, 11 patients died; of those, three deaths were directly associated with anticoagulation therapy and there was one death of unknown cause. During follow-up, the quality of OAC control was significantly better in the SMG than in the RCG. In elderly patients, long-term self-management of oral anticoagulation is superior for the prevention of major thromboembolic and bleeding complications and for the quality of oral anticoagulation control compared to routine care for a mean follow-up period of three years.