Using data from smokers (N = 591) who enrolled in an 8-week smoking cessation program and were then followed for 15 months, the authors tested the thesis that self-efficacy guides the decision to initiate smoking cessation but that satisfaction with the outcomes afforded by quitting guides the decision to maintain cessation. Measures of self-efficacy and satisfaction assessed at the end of the program, 2 months, and 9 months were used to predict quit status at 2, 9, and 15 months, respectively. At each point, participants were categorized as either initiators or maintainers on the basis of their pattern of cessation behavior. Across time, self-efficacy predicted future quit status for initiators, whereas satisfaction generally predicted future quit status for maintainers. Implications for models of behavior change and behavioral interventions are discussed.