Of 56 male smokers, 34 were randomly assigned (by 60% random odds) to quit smoking immediately, whereas the remaining 22 were assigned to quit after an additional 31 days. Compensation ($300) was contingent on abstinence for a minimum of 31 or 2 days (depending on random assignment) and completion of all experimental sessions. Contingencies for the immediate-quit group required 31 days of abstinence; those for the delayed-quit group required only 2 days of abstinence. Contingency duration (31 vs. 2 days) predicted days to relapse. All but 4 of the 31-day contingency participants maintained abstinence for at least 31 days, whereas only 3 of the 2-day contingency group abstained for 31+ days. However, 31-day contingencies did not result in longer postcontingency time to relapse. Higher trait neuroticism, depression, and psychopathic deviate scores predicted decreased time to relapse. Prequit cotinine concentrations and Fagerström Tolerance Questionnaire scores failed to predict time to relapse.