Although a variety of psychological benefits have been attributed to regular exercise, few experimentally controlled studies of healthy individuals currently exist. One hundred twenty healthy, sedentary, middle-aged men and women were randomly assigned to either a 6-month home-based aerobic exercise training program or to an assessment-only control condition. Adherence across the 6-month period was found by both self-report and heart rate microprocessor methods to exceed 75% in both sexes. To assess changes in a variety of psychological variables over time, a 14-item Likert rating scale was completed and returned on a biweekly basis throughout the 6-month period. Slope analyses conducted on the 11 items attaining acceptable test-retest reliability coefficients showed significant between-groups differences on the 3 items most closely associated with the actual physical changes that occurred with exercise (all ps less than .004). Implications in relation to repeated measurement of psychological changes in nonclinical populations and the determination of the relevant population-, activity-, and program-specific parameters involved are discussed.