To examine the effectiveness of self-debriefing as compared to instructor debriefing in the change of nontechnical skills performance of anesthesiology residents. Prospective, randomized, controlled study. A university hospital simulation center. : Fifty anesthesiology residents. Subjects were instructed in the principles of nontechnical skills for crisis management. Subsequently, each resident participated in a high-fidelity simulated anesthesia crisis scenario (pretest). Participants were randomized to either a video-assisted self-debriefing or instructor debriefing. In the self-debriefing group, subjects reviewed their pretest scenario by themselves, guided by the Anesthetists' Non-Technical Skills scale. The instructor debriefing group reviewed their pretest scenario guided by an expert instructor also using the Anesthetists' Non-Technical Skills scale as a framework. Immediately following their respective debriefings, subjects managed a second simulated crisis (post-test). After all data were collected, two blinded experts independently rated videos of all performances in a random order using the Anesthetists' Non-Technical Skills scale. Performance significantly improved from pretest to post-test (p < .01) regardless of the type of debriefing received. There was no significant difference in the degree of improvement between self-debriefing and instructor debriefing (p = .58). Nontechnical skills for crisis resource management improved with training, as measured by the Anesthetists' Non-Technical Skills scale. Crisis resource management can be taught, with measurable improvements. Effective teaching of nontechnical skills can be achieved through formative self-assessment even when instructors are not available.