To examine person-specific urge-to-smoke trajectories during the first 7 days of abstinence and the relationship of trajectory parameters to continuous abstinence, demographics, medication and smoking history. Hierarchical linear modeling was used to model person-specific trajectories for urge-to-smoke. Two university-based smoking cessation trials. Treatment-seeking smokers in a clinical trial of transdermal nicotine (n = 275) versus nicotine spray (n = 239) and of bupropion (n = 223) versus placebo (n = 198). Self-reported urge-to-smoke for 7 days after the planned quit date, and 7-day point prevalence and continuous abstinence at end of treatment (EOT) and 6 months. Urge-to-smoke trajectory parameters (average level, slope, curvature and volatility) varied substantially among individuals, had modest intercorrelations and predicted continuous and point prevalence abstinence at EOT and at 6 months. Higher trajectory level, slope and volatility were all significantly (P ≤ 0.001) associated with a reduced likelihood of abstinence at EOT (odds ratios 0.44-0.75) and at 6-month follow-up (odds ratios from 0.63 to 0.78), controlling for demographic, medication and smoking use variables. Higher urge-to-smoke trajectory parameters of level, slope and volatility (measured over 7 days) predict continuous and 7-day point prevalence at EOT and 6 months. Although there were some associations of trajectory parameters with demographics and smoking history, the associations of trajectory parameters with relapse were relatively uninfluenced by demographics and smoking history.