We studied relationships among key elements of selected theoretical models of adult learning applied to CME, using a descriptive survey technique and nonparametric statistical analysis. Questionnaires were administered on-site and three months after the course to 346 physicians who attended a seven-day National Emergency Medicine CME program in October 1991; response rates were 91 percent and 49 percent respectively. The CME program's content was based on the core content curriculum developed by the Society of Academic Emergency Medicine (SAEM), the American Board of Emergency Medicine (ABEM), and the American College of Emergency Physicians (ACEP). Course faculty members assisted attendees in self-selection of learning projects and instructional resources to meet their specific self-defined learning styles and needs. We measured program impact at three levels: (1) perception (program quality and effectiveness), (2) competency (the ABEM scores of those who took the written boards), and (3) change in physician behavior (self-reports on incorporation of content into their practices). We also assessed the effectiveness of various instructional methods. The overall quality rating was 4.78 (scale 5=hight 1=low), and the overall effectiveness rating was 3.50 (scale 4=high, 1=low). Recent data support a moderate level of validity in predicting performance/competency in practice from performance on written multiple-choice questions (MCQ's). Of attendees who were nonresidency trained and reported written ABEM scores, 80 percent attained the passing level—considered “passing”—of 75 percent At three months follow-up, 46 percent of respondents stated they had incorporated knowledge/skills obtained from the program into their practices. An incidental finding was a statistically significant greater use of computers by those who had been graduated from medical schools after 1980 (p = 0.026). Physicians as adult learners are self directedlexperientially oriented; influenced by their physical agelstage of life; and must “see and feel” their learning needs. These results suggest that this program is effective in helping physicians because of the application of adult learning principles. Specifically, the addition of a nontraditional CME portion onto a traditional program found considerable favor with the participants.