Regular physical activity (PA) provides health benefits; however, at least 60% of the population fails to engage in the recommended amount of PA required to produce these health benefits. The primary purpose of the study was to examine over a 12-week structured exercise program the effect of a multifaceted efficacy intervention (MEI-i.e., task and specific types of self-regulatory efficacy) on objectively measured exercise behavior. Secondary purposes were to examine the effect of the MEI on both task and self-regulatory efficacy levels; and to determine whether efficacy beliefs could predict exercise behavior. Relatives of colon cancer patients (N = 140) were enrolled in an exercise program, and were randomized to either a MEI or attention control condition, and took part in classroom sessions. Behavior was assessed throughout the 12-week program using objective measures of frequency, duration, and intensity of exercise, and dropout rates, while self-reported task, barrier, scheduling, goal-setting, and relapse prevention efficacy were assessed at baseline and weeks4, 8, and 12. The MEI group exercised for longer duration in the early phase of the program (i.e., 0-4weeks); however, no significant differences were noted for exercise frequency and intensity. Differential dropout was found favoring the MEI group at weeks8 and 12. No treatment condition differences were found for reported efficacy beliefs. Proceeding self-efficacious beliefs were associated with objective measures of behavior. A MEI grounded in Social Cognitive Theory was partially effective in influencing colon cancer relatives exercise behavior.