Bariatric surgery has emerged as an important tool in the fight against morbid obesity. However, reviewers have noted that there is a scarcity of long-term clinical surveillance data for bariatric surgery beyond 1-year follow-up and that a high percentage of patients are lost to follow-up, raising questions regarding the accuracy of current outcomes estimates. A meta-analysis of clinical reports providing bariatric surgery weight loss outcomes for morbidly obese patients was conducted over the period 2003-2007. Studies included were randomized controlled trials, nonrandomized controlled trials, and consecutive case series involving patients receiving either laparoscopic adjustable gastric banding (LAGB) or laparoscopic gastric bypass (LGB) surgery. Included studies involved n = 7,383 patients and were largely academic hospital-based (78.6%) and retrospective in design (71.4%). Weight loss outcome was defined by percent excess weight loss (%EWL). Composite estimates showed a significantly greater %EWL for LGB surgery (62.6%) compared to LAGB (49.4%). The superiority of LGB persisted at all three postsurgical time points examined (1, 2, and >3 years). Problems were identified regarding incomplete or suboptimal data reporting in many studies reviewed, and high patient attrition was evident at 2-year (49.8% LAGB, 75.2% LGB) and >3-year (82.6% LAGB, 89% LGB) end points. This meta-analysis confirms the superiority of LGB to LAGB in %EWL found in earlier studies. Although problems in study quality raised significant concerns regarding the validity of current weight loss estimates in this area, there was no evidence of publication bias.