PURPOSE: To systematically review the published literature on graduate medical education (GME) curricula in clinical epidemiology, critical appraisal, and evidence-based medicine (EBM). METHOD: The author searched the Medline and Educational Resources Information Center (ERIC) databases from 1973 through 1998, and also searched the references of the captured papers. The author reviewed all peer-reviewed reports of GME curricula (with or without effectiveness studies) in critical appraisal, clinical epidemiology, or evidence-based medicine, extracting objectives, formats, and evaluations (including effectiveness, process, and satisfaction). For effectiveness evaluations, he also identified the outcomes, outcome measures, methodologic characteristics, and results. RESULTS: The search produced 18 reports. The most common objective of the curricula described in the reports was improving critical skills; the most common format was resident-directed small-group seminar. The most common outcome-evaluation measure was a multiple-choice examination. Only seven of the reports evaluated the curricula's effectiveness, and only four met a minimum methodologic standard of a pretest-posttest controlled trial. The impacts on critical appraisal skills of the curricula in those four reports ranged from no effect to a 23% net absolute increase in test scores. CONCLUSION: These reports provide useful guides for medical educators, but many suffered from incomplete descriptions and inadequate evaluations of their curricula. The curricula themselves often focused on critical appraisal to the exclusion of other EBM skills and had limited effectiveness. In addition to increased methodologic rigor, future studies should focus on more meaningful outcome evaluations. Curricula should use residents' actual clinical experiences and teach EBM skills in real time in existing clinical and educational venues.