BACKGROUND: As opportunities for quality improvement become more visible, educational planners, health services researchers, and policymakers search for strategies that lead to better clinical outcomes. Continuing medical education (CME) is one such strategy, but the impact of CME is poorly defined in relation to clinical outcomes, and efforts to standardize definitions of clinical outcomes are in varied stages of development. METHODS: The Johns Hopkins University Evidence-based Practice Center conducted a systematic review of the effectiveness of CME for the Agency for Health Care Research and Quality. From the review, 37 studies were used by the guideline panel to answer questions about improvement in clinical outcomes. Recommendations were made using the American College of Chest Physicians guideline grading system. RESULTS: Multiple media, multiple techniques of instruction, and multiple exposures to content are suggested to meet instructional objectives intended to improve clinical outcomes. CONCLUSIONS: There are models to describe and guide the planning and evaluation of CME, and there are models to measure quality of care. Research and practice of CME must be defined in relation to guideline implementation and quality improvement and other interventions and systems intended to improve or measure clinical outcomes. Further research is required to identify the qualities essential for measuring causal linkages thought to exist among CME, physician behavior, and clinical outcomes.