We investigated the relationship between smoking status and depression symptom score in a large prospective cohort of women (n = 7,089) followed at regular intervals during and immediately following pregnancy. We wished to determine whether depression symptom score predicts subsequent likelihood of failed or successful smoking cessation attempts, and whether smoking cessation during and immediately following pregnancy is associated with change in depression symptom score. Longitudinal latent class analysis was used to empirically derive smoking trajectories. These latent smoking classes were characterized using adversity measures collected at baseline. Depression symptom score at baseline was used to attempt to predict class membership. The mean depression symptom score and change in depression symptom score was calculated within each smoking trajectory for each time point. We identified seven distinct smoking trajectories, with five reflecting a transition from smoking to nonsmoking, and in four of these a relapse back to smoking. Depression symptom score at baseline did not strongly differentiate between smoking trajectories. Those that resulted in abstinence demonstrated the lowest depression symptom scores at that time point. The analysis of change in depression symptom score suggested different relationships between depression symptom score, smoking cessation, and relapse among the various trajectories. Our data suggest the relationship between depression symptom score and smoking status may differ across the trajectories we identified. In general, smoking cessation appears to be associated with a reduction in depression symptom score. Future studies should explore this possibility in more detail, including whether these relationships differ in clinical and nonclinical samples.