The ability to cope effectively during high-risk situations (e.g., exposure to drug-related stimuli during acute withdrawal) is essential for forestalling relapse during attempts to quit problematic substance use. Attempting to exert executive cognitive control over affective reactions is a frequently employed strategy for managing temptation and sustaining cessation. Such attempts are not failsafe, however, with many individuals succumbing to temptation despite reporting the use of cognitive coping strategies. The reasons for such failure, as well as for the observation that the efficacy of coping varies significantly both within and between individuals, remain largely unknown. The goal of the present study was to address this important knowledge gap by investigating the mechanisms underlying cognitive coping in cigarette smokers, with two specific aims. The first aim was to examine the neural correlates of the use of two different forms of cognitive coping during drug cue exposure, with the prediction that the use of a non-self-referential strategy would be associated with relatively greater activation of the DLPFC than a strategy that entails the use of self-referential information. In contrast, it was hypothesized that a strategy that involves the generation and maintenance of self-relevant information would be associated with comparatively greater activation of portions of the anterior medial prefrontal cortex than a strategy in which the focus is on non-self-referential information. The second aim of the study was to examine whether non-self-referential and self-referential coping strategies are differentially moderated by individual differences in working memory capacity, with the hypothesis that working memory ability would more strongly predict the magnitude of cue-elicited activation of the DLPFC during the use of a non-self-referential coping strategy than during the use of a self-referential coping strategy. Findings suggest non-self-referential and self-referential coping indeed are associated with different patterns of neural activation during cue exposure, although the specific relationships that were observed proved to be more complex than initially hypothesized. In contrast to expectations, however, working memory capacity did not differentially moderate the activation of the DLPFC and measures of cue-reactivity. Potential implications and extensions of these findings are discussed.