This study examined predictors of exercise maintenance following completion of a physical activity intervention. Sedentary adults recruited through newspaper advertisements were randomly assigned to receive either (a) a motivation-matched intervention with feedback reports that were individually tailored (IT) to psychological variables from social cognitive theory and the Transtheoretical Model via computer expert system, or (b) a standard, print-based intervention (ST). The intervention phase of the study included mailed assessments and intervention materials at baseline, 1, 3, and 6 months. An assessment-only follow-up was conducted 6 months after the end of the intervention (Month 12). Participants were assessed for current physical activity participation, motivational readiness for physical activity, a number of psychological constructs posited to influence participation in physical activity (e.g., self-efficacy), and current affect. Significantly more participants in the IT condition met or exceeded exercise participation goals at the end of the intervention period and maintained this level of physical activity through the Month 12 follow-up compared to ST participants. Prospective analyses revealed significant differences in several psychological constructs both at program entry (baseline) and the end of the intervention period between individuals who maintained their physical activity participation through Month 12 and those who did not. Results suggest that the maintenance of physical activity following the end of an active intervention program may be influenced by attitudes and behaviors acquired along with increased participation in physical activity, as well as by preexisting characteristics that individuals bring into treatment.