Introduction: the probability of recurrence is present in near of 80% of the patients with upper non-variceal digestive bleeding. Objective: to determine the rebleeding risk applying the complete Rockall score in patients with upper non-variceal gastrointestinal bleeding. Material and Methods: a retrospective analysis was carried out in 118 patients admitted to the emergency room and endoscopy unit of “Dr. Luis Diaz Soto” hospital with non-variceal upper gastrointestinal bleeding between September 2011 and March 2013.The Rockall scores were calculated and patients were divided into three risk categories: low risk, moderate risk and high risk. Receiver operating characteristic (ROC) curves were plotted for rebleeding. The discriminative accuracy of the risk scores was assessed by the area under the ROC curve. Results: high clinical Rockall scores >5 were associated with rebleeding. The accuracy of the complete Rockall score in predicting rebleeding was good (area under ROC: 0,821, IC 95%: 0,664-0,977). The sensitivity was of a 63 % and the specificity of 11, 8 %, j= 0, 53. Conclusions: the Rockall score is clinically useful, rapid and accurate in predicting rebleeding, by what it can be entered in the clinical practice of this hospital.