Background Patients with gastrointestinal bleeding during anticoagulant and/or antiplatelet therapy represent a clinical challenge. Aim To determine the risk/rates of rebleeding, vascular events and death in patients treated with antiplatelet or anticoagulant agents who developed major gastrointestinal bleeding Methods This was an observational cohort study of patients who developed gastrointestinal bleeding while on antiplatelet and/or anticoagulant therapy. Drug use information was collected prospectively during bleeding events. Cox proportional hazards models were used to evaluate rebleeding, vascular events and death. Results Among 871 patients (mean age 78.9 +/- 8.6 years), 38.9% used an anticoagulant, 52.5% used an antiplatelet and 8.6% used both; 93.1% interrupted treatment after gastrointestinal bleeding and 80.5% restarted therapy within 7.6 +/- 36.4 days; 38.7% had upper gastrointestinal bleeds, 46.7% lower gastrointestinal bleeds and 14.6% gastrointestinal bleeds of unknown origin. Median follow-up was 24.9 months (IQR: 7.0-38.0). Resumption of both therapies was associated with a higher risk of rebleeding, lower risk of ischaemic events or death and a similar risk for upper and lower gastrointestinal events. Resumption of therapy <= 7 days after bleeding showed a similar pattern with no differences in death. Rebleeding rates were higher in anticoagulant vs antiplatelet patients (138.0 vs 99.0 events per 1000 patient-years), and the bleeding location was identical in 61.8% of cases. Conclusions Resumption of anticoagulant or antiplatelet therapy after a gastrointestinal bleeding event was associated with a lower risk of vascular events and death and a higher rebleeding risk. The benefits of early reinstitution of anticoagulant/antiplatelet therapy outweigh the gastrointestinal-related risks.