RATIONALE: Tobacco has a higher rate of dependence than other drugs of abuse. However, the psychopharmacological effects of nicotine are incongruent with the tenacity of tobacco addiction since nicotine does not produce robust euphoria in humans or self-administration in rodents. A potential explanation is that nicotine amplifies the salience of other stimuli that have some incentive value, which could influence the initiation and persistence of smoking. However, the neural mechanisms of this process are unknown. OBJECTIVES: One way that nicotine may amplify the salience of other stimuli is by enhancing reward prediction errors. We hypothesized that nicotine would enhance the neural response to unexpected (relative to expected) rewards compared to placebo. METHODS: Twenty-three nonsmokers underwent two fMRI scans, following nicotine (1 mg) or placebo administration, while performing an outcome expectation task. In the task, a pair of cues was associated with either a subsequent reward (the image of a $100 bill) or a nonreward (the image of a blurry rectangle). On 20% of trials, the cue was followed by an unexpected outcome. RESULTS: Although nicotine did not affect the magnitude of prediction errors relative to placebo, nicotine did increase BOLD activation in the anterior insula/inferior frontal gyrus and decrease activation in the caudate across all outcome types (including both rewards and nonrewards). CONCLUSIONS: The insula and caudate could play a role in the initial effects of nicotine in nonsmokers, and these changes in baseline may be the mechanism that underlies how nicotine amplifies the salience of nondrug stimuli.