Abstract Background Currently, no clinically useful scoring system is available to stratify the mortality risk for patients undergoing gastric bypass (GBP). We propose the obesity surgery mortality risk score as a clinically useful score system to predict the mortality risk for patients undergoing GBP. Methods Prospectively collected data from 2075 consecutive patients undergoing GBP at a single university from 1995 to 2004 were analyzed to determine the preoperative factors correlating with 90-day mortality. Results Four independent variables correlated with mortality using multivariate analysis, including body mass index ≥50 kg/m 2 (odds ratio [OR] 3.60, 95% confidence interval [CI] 1.44–8.99), male gender (OR 2.80, 95% CI 1.32–5.92), hypertension (OR 2.78, 95% CI 1.11–7.00), and a novel variable pulmonary embolus risk, that included previous thrombosis, pulmonary embolus, inferior vena cava filter, right heart failure, and obesity hypoventilation (OR 2.62, 95% CI 1.12–6.12). A fifth variable, patient age ≥45 years (OR 1.64, 95% CI 0.78–3.48), significant on univariate analysis, was added to the ultimate scoring system because of its significance in other studies. A scoring system was developed by arbitrarily scoring the presence of each independent variable as equal to 1 point, resulting in an overall score of 0–5 points for each patient. The factors were grouped into 3 risk classes (A, B, or C) to increase the evaluable cases in each class (e.g., <1% of 2075 patients accrued all 5 points). The mortality rate among the 3 risk classes was significantly different: class A, 0.31%; class B, 1.90%; and class C, 7.56%. Conclusion The analysis reveals that mortality risk for gastric bypass can be stratified based upon independent variables that can be identified before surgery. The OS-MRS, a simple, clinically relevant scoring system, is proposed, which stratifies mortality risk into low (Class A), intermediate (Class B), and high (Class C) risk groups in the current study population. This risk assessment scoring system may contribute to surgical decision making in bariatric surgery if its ability to stratify risk is validated in subsequent studies.