There is currently a need in the health care industry to develop and implement strategies to improve both the quality and cost effectiveness of health care delivery. Continuing medical education (CME), like other components of the health care system, is being pressured for more accountability in physician education programs. Specifically, CME professionals are being asked to address quality assurance issues (either physician performance or health care outcomes) as measures of the impact of educational programs and activities. Major research studies reported in the CME literature have pointed to the importance of determination of practice or real learning needs as a prerequisite for effective education. Furthermore, it has been suggested that research in CME should focus on evaluating the effects of CME interventions on major causes of morbidity and mortality. The funded, on-going project described in this paper is a continuing medical education efsort designed to address and correct dejiciencies in clinical quality and cost effectiveness documented by a mandatory statewide hospital data collection system. The database measures quality and effectiveness of care using two specific data elements: severity of illness upon admission and patient outcome upon discharge, in terms of morbidity and mortality for selected Diagnostic Related Groups. Key clinical findings are the focus. This article provides a conceptual framework and descriptive analysis of (1) the rationale for using large administrative databases in designing, implementing, and evaluating continuing medical education programs, and (2) the replicable problem-solving process that can be used in identifying the real learning needs of physicians from large administrative databases.