Beyond their incentive value, visual sexual stimuli are thought to have intrinsically rewarding properties that may contribute to the rising prevalence of problematic pornography use. However, whether excessive consumption of visual sexual stimuli fits classic models of addiction and involves reinforcement-based learning remains controversial. To address this question, the present study focused on the interplay of individual differences in trait sexual desire (specifically, the drive to engage in solitary sexuality) with stimulus modality in appetitive Pavlovian conditioning. 62 heterosexual participants (final sample, 36 women) underwent two sessions of differential conditioning, spaced one week apart. During one learning session, neutral cues were reinforced (50 %) by presentation of visual sexual stimuli, while auditory sexual stimuli served as unconditioned stimuli during the other session. Indexing both sexual arousal and appetitive learning, pupil dilation (as well as startle modulation) was used to track the acquisition of conditioned responses. Results revealed that solitary sexuality was associated with blunted differential pupillary responses to cues predicting visual (yet not auditory) sexual stimuli and less sensitization across trials, presumably reflecting reduced anticipatory arousal (consistent with self-report findings) and/or altered processing of uncertainty. At the same time, both enhanced startle habituation and valence ratings suggest that the preference for erotica was unaffected in individuals high in solitary sexuality. Fitted computational models provide additional evidence for a link to divergent learning trajectories. Taken together, our findings underscore the special nature of visual sexual stimuli (compared to auditory sexual stimuli) and support the view that excessive consumption may reflect a dispositional reward deficiency that drives individuals to seek out more intense stimulation.