Cigarette smoking remains the leading cause of morbidity and mortality in the U.S., and smoking rates are high. Nicotine Replacement Therapy (NRT) is sold over-the-counter and often distributed in smoking cessation clinical trials. With proper use, the nicotine patch is effective at helping individuals quit smoking and stay quit. However, rates of NRT adherence are low, and there is little research on the psychological predictors of compliance. Drawing from the larger medication adherence literature, we know that depressive symptoms predict poor medication adherence, though this has never been studied as it relates to NRT use. Furthermore, Behavioral Activation Treatment for Smoking (BATS) is hypothesized to reduce depressive symptoms through increases in both positive reinforcement from the environment and goal-directed activation. Thus, it is possible that individuals who receive this treatment may exhibit increased compliance to the patch first through increases in environmental rewards or goal-directed activation, and subsequently though decreases in depressive symptoms. Using data from a stage-II randomized controlled trial (RCT) examining BATS compared to standard smoking cessation treatment (ST), we utilized a serial mediation model to examine this question. The aims of the study were: (1) to examine a serial mediation model in which BATS would affect NRT adherence through increases in environmental reward at mid-treatment and decreases in end-of-treatment depressive symptoms; and (2) to examine a serial mediation model in which BATS would affect NRT adherence through increases in goal-directed activation mid-treatment and decreases in end-of-treatment depressive symptoms. We hypothesized that BATS would produce increases in environmental reward, which in turn would decrease depressive symptoms, which in turn would predict greater patch use. Our second model examined goal-directed activation as a mediator, and we hypothesized that BATS would predict increases in goal-directed activation, which in turn would predict decreases in depressive symptoms, which in turn would predict increases in NRT adherence. We examined these models using Structural Equation Modeling (SEM). Our serial mediation model examining treatment conditionenvironmental rewardsdepressive symptomsNRT adherence was not supported, and neither was our model examining treatment conditiongoal-directed activationdepressive symptomsNRT adherence. However, we found significant pathways from mid-treatment goal-directed activation and NRT adherence, and end-of-treatment depressive symptoms to NRT adherence in our model examining goal-directed activation. Changes in both goal-directed activation and depressive symptoms may be key when predicting NRT adherence in a sample of adult smokers enrolled in smoking cessation treatment.