OBJECTIVE:To evaluate evidence in recent authoritative 'Evidence-Based Medicine' (EBM) reports on surgery for severe obesity. METHODS: Focused review of Index Medicus citations and authors' own databases of publications on surgery for obesity, . RESULTS: EBM criteria for assessment of strength of evidence requiring randomized controlled studies (RCTs) in these reports are inappropriate for evaluating invasive treatments such as surgery, which have robust physiological effects, are difficult to reverse and may have more serious side effects than the drug studies for which the criteria were promulgated. Flaws in these reports include omissions of important studies demonstrating improvements in comorbidity, factual errors in descriptions of operations and faulty analyses of outcomes of laparoscopic approaches. There are misinterpretations of cited papers, and inclusion of obsolete operations as well as a study generated during the 'learning curve' of an avowed complex procedure. CONCLUSION: EBM analyses of surgical modalities affecting access to care require relevant evaluation criteria, true peer review and expert consultation. Authors' claims of objectivity by invoking use of evidence-based criteria applicable to drug treatment and other easily reversible forms of therapy are questionable. Decisions based on flawed EBM reports may adversely affect access to care for millions of severely obese patients.