The COPD Patient Management European Trial (COMET) investigated the efficacy and safety of a home-based COPD disease management intervention for severe COPD patients.The study was an international open-design clinical trial in COPD patients (forced expiratory volume in 1 s <50% of predicted value) randomised 1:1 to the disease management intervention or to the usual management practices at the study centre. The disease management intervention included a self-management programme, home telemonitoring, care coordination and medical management. The primary end-point was the number of unplanned all-cause hospitalisation days in the intention-to-treat (ITT) population. Secondary end-points included acute care hospitalisation days, BODE (body mass index, airflow obstruction, dyspnoea and exercise) index and exacerbations. Safety end-points included adverse events and deaths.For the 157 (disease management) and 162 (usual management) patients eligible for ITT analyses, all-cause hospitalisation days per year (mean±sd) were 17.4±35.4 and 22.6±41.8, respectively (mean difference −5.3, 95% CI −13.7 to −3.1; p=0.16). The disease management group had fewer per-protocol acute care hospitalisation days per year (p=0.047), a lower BODE index (p=0.01) and a lower mortality rate (1.9% versus 14.2%; p<0.001), with no difference in exacerbation frequency. Patient profiles and hospitalisation practices varied substantially across countries.The COMET disease management intervention did not significantly reduce unplanned all-cause hospitalisation days, but reduced acute care hospitalisation days and mortality in severe COPD patients.