This study examines two patient education interventions during Phase I cardiac rehabilitation. Knowles' adult instructional theory provided theoretical guidance for the experimental intervention. In addition, Bandura's self-efficacy theory influenced the use of experiences designed to enhance patients' confidence in their ability to engage in recovery behaviors. The experimental intervention was structured to promote active patient participation including problem solving and immediate application of program content. The contrasting educational intervention emphasized verbal persuasion and used a structured slide/lecture format where predetermined information was given to patients. Both interventions were conducted in group settings and followed the same exercise protocols. Forty randomly assigned, Phase I coronary artery bypass graft patients in each of the interventions completed a minimum of three educational sessions and were included in data analysis. The ability to rehearse behaviors that patients determined to be important, both by watching others and through active performance was shown to significantly impact patients' confidence in their ability to carry out four of five typical recovery behaviors taught in a Phase I cardiac program. Self-reports of performance for the same recovery behaviors correlated significantly with estimates of self-efficacy, indicating that the educational programs may also impact behavior.