Numerous studies have reported that exercise is safe and beneficial for breast cancer survivors; however, long-term adherence to exercise programs is not easy to accomplish. This secondary analysis examined the demographic and clinical characteristics, adherence to exercise, and cognitive-behavioral intervention components data collected on 120 postmenopausal women with a history of breast cancer and bone loss who had been randomized to the exercise group in a 24-month study. Hierarchical regression was used to identify variables that predicted adherence to exercise. Mean adherence to exercises was 61.89%. Feedback and support were the most frequently used cognitive-behavioral intervention components. In hierarchical regression, predictors for adherence to exercise were feedback (β = .40, p < .001) and adherence to exercise in the previous time period (β = .31, p < .001). Participants receiving more frequent feedback had higher adherence to exercise.