A randomized prospective evaluation of the gastric and jejunoileal bypass procedures for morbid obesity was performed. The gastric bypass was performed predominantly as a 90% gastric exclusion with a Roux-en-Y reconstitution. The jejunoileal bypass was an end-to-end anastomosis between 30 cm of jejunum and 25 cm of terminal ileum, the bypassed segment of small bowel being decompressed by an end-to-side ileocolostomy. There were 32 patients in the gastric group and 27 in the jejunoileal group. The two groups were comparable in age, preoperative weight and height. There were no postoperative deaths, but the gastric bypass operation was associated with a slightly higher early complication rate indicating it is a more technically demanding procedure. Late sequellae were more prominent in the jejunoileal bypass group and included significant diarrhea in 56% and need for medication in 74%. Kidney stones and cholelithiasis also complicated the jejunoileal group and were not seen after gastric bypass. All patients showed fatty metamorphosis on the original liver biopsy. This had worsened in 75% of the jejunoileal group at one year whereas it had improved or was stable in all of the patients in the gastric group.