Attentional bias to drug-related cues is widely considered a core feature of addiction, yet its empirical detection is highly sensitive to task design and stimulus selection. In this study, we used a computer mouse-tracking paradigm to examine attentional bias toward cigarette-related, pleasant, unpleasant, and neutral images in 96 participants (49 smokers, 47 nonsmokers). Trajectory deviation, measured via maximum absolute deviation (MAD), served as an index of attentional bias. Both groups showed stronger bias for emotionally salient distractors, especially pleasant images, highlighting how the inclusion of emotionally matched control stimuli could reveal overlooked confounds in addiction research. Contrary to expectations, smokers did not exhibit greater attentional bias to cigarette cues than nonsmokers. Prior work has identified individual differences in cue reactivity among smokers. To investigate whether similar differences were present here, we applied k-means clustering to the MAD data. This analysis revealed two subgroups of smokers: one selectively biased toward cigarette cues, and another more reactive to pleasant stimuli. We replicated these results in an independent sample of 85 smokers. This dissociation aligns with prior neuroimaging findings, and suggests that the mouse-tracking paradigm may serve as a scalable, cost-effective tool for detecting reliable individual differences in reactivity to drug-related and non-drug-related pleasant stimuli.