Sixteen adults with traumatic brain injuries participated in an eight-week goal setting training program designed to assess the effects of direct involvement in goal-setting on treatment outcome. Participants were randomly assigned to either a condition designed to facilitate high involvement (HI) in goal setting or to a low involvement (LI) condition. From pretesting to posttesting, both HI and LI participants significantly improved in attaining their goals. However, at the two-month follow-up, HI participants maintained gains from pretesting, whereas LI participants regressed to their pretesting level.Goal setting and goal attainment have been of considerable interest to professionals working within the field of rehabilitation. A goal can be described as the end result of a course of action. Goal setting typically includes three major components: (a) a precise description of the desired performance, (b) the conditions under which the behavior will occur, and (c) the time frame for assessing the target behavior (Zahara & Cuvo, 1984).Psychologists and allied health professionals working with persons with chronic disabilities have emphasized the importance for the client to be directly involved in goal setting. According to Trieschmann (1974), if persons with disabilities do not set goals for themselves, goals may be set for them, resulting in less effective overall rehabilitation. Lloyd (1986) suggested that directly involving a client in the goal setting process is self-reinforcing. LaFerriere and Calsyn (1978) have asserted that adults with disabilities who are directly involved in goal setting are more likely to report greater motivation for change. Furthermore, active involvement of clients may facilitate greater levels of goal attainment as a result of a mutual refining of treatment goals (France, 1985;Wilier & Miller, 1976).