Background: Gastric bypass (GBP) is the most effective method for controlling morbid obesity. Previously we showed that extending the length of the Roux limb increased weight loss. We have done over 400 obesity operations during the past 20 years. The current study consists of patients from the last 10 years of our experience and compares short to extended Roux-en-Y GBP. Methods: Data from all patients operated at the Ottawa General Obesity Clinic were entered into a database on an ongoing basis, and those from the past 10 year were analyzed. All patients had standardized preoperative investigations and postoperative follow-up. Details of these and of the operative technique are provided in the manuscript. Results: The preoperative weight and BMI were 129 ±2 kg, and 46 ± 2 kg/m², respectively. The mean weight loss prior to surgery was −2± 21 kg. The results were classified, by percentage weight loss as: ‘excellent’ = > 35%; ‘good’ = 25-34%; ‘poor’ = 15-24%; and ‘failure’ = < 15%. Sixty-five patients (69%) were available for 2-year follow-up. At this time mean percentage weight loss was 34 ± 2 versus 40 ± 1 for the short Roux (45-135cm) and long Roux (180-225 cm) groups, respectively (P<0.01). There were no deaths, leaks, splenectomies or intraabdominal infections. The incidence of hernia and/or reoperation for bowel obstruction was 35/121 or 29%. The overall incidence of diarrhea was 16/121 (13%) and 6/121 (5%) at 12 and 24 months. Quality of life is significantly impaired in at least three of these patients, all with extended limbs. Major vitamin deficiencies, alterations in liver functions, or other metabolic complications did not occur. Conclusions: Gastric bypass is the procedure of choice for morbid obesity. Weight loss is marginally improved in proportion to the length of the Roux limb but at a risk of diarrhea, which occasionally may not manifest itself for 8 to 12 months. It is important that methods be devised to correct followup, incisional hernias and diarrhea.