Objective: The present study examined whether dynamic day-to-day variations in self-efficacy predicted success in quit attempts among daily smoking adolescents. Design: A sample of 149 adolescents recorded their smoking and self-efficacy three times per day during 1 week prior to and 3 weeks after a quit attempt. Main Outcome Measures: The first lapse, second lapse, and relapse after at least 24 hours of abstinence from smoking were the main outcome measures. Results: Self-efficacy was relatively high and moderately variable prior to the first lapse, but decreased and became more variable thereafter. Lower self-efficacy as measured at the lapse assessment significantly increased the risk that a second lapse and relapse would occur. Individual differences in baseline self-efficacy did not predict any of the treatment outcomes. The time-varying analyses, however, showed that lower self-efficacy on a given day predicted the first lapse, the second lapse, and relapse on the succeeding day. Daily concomitant smoking (any smoking on the preceding day) was not significantly related to relapse. Conclusion: The present results emphasize the importance of self-efficacy among adolescents in cessation and highlight the need for dynamic formulations and assessments of adolescents' self-efficacy and relapse.