Objective To compare the outcomes of patients with non-variceal upper gastrointestinal bleeding (NVUGIB) taking aspirin for primary prophylaxis to those not taking it.Methods Patients without any vascular disease (coronary artery or cerebrovascular disease) who were admitted to the 89th Hospital of PLA between 2006 and 2016 with non-variceal upper gastrointestinal bleeding were included.The frequencies of in-hospital mortality,re-bleeding,severe bleeding,need for surgery or embolization,and of a composite outcome of the 4 bleeding related adverse outcomes were compared between patients receiving aspirin and those on no antithrombotics.We also compared the frequency of in hospital complications and length of hospital stay between the two groups.Results Of the 179 eligible patients,47 were on aspirin and 132 patients were on no antithrombotics (control group).Patients in the aspirin group (58.4 years) were elder than that in controls (66.8 years) (P ＜0.01).Patients in the aspirin group had significantly more co-morbidities,including diabetes mellitus [25.5 ％ (12/47) vs 12.1％ (16/132)] and hypertension [46.8％ (22/47) vs 28.0％ (37/132)],respectively,as well as dyslipidemia [21.3％ (10/47) vs 6.1％ (8/132)] (P ＜ 0.01).Smoking was more frequent in the aspirin group [36.2％ (17/47) vs 22.7％ (30/132),P ＜ 0.05].The frequencies of endoscopic therapy and surgery were similar in both groups,while the frequencies of in-hospital complications and re-bleeding were similar in the two groups,too.However,the patients who were on aspirin had lower in hospital mortality rates (2.1％ vs 13.6％,P ＜0.01),shorter hospital stay (4.9 d vs 7.0 d,P ＜0.05),and fewer composite outcomes (10.6％ vs 24.2％,P ＜ 0.05).Conclusions Patients who present with NVUGIB while receiving aspirin for primary prophylaxis had fewer adverse outcomes.Thus aspirin may have a protective effect beyond its cardiovascular benefits.