Introduction: Understanding of the cognitive processes, and in particular social information processing bias, that generate cognitive distortions among sexual offenders with a learning disability has been hampered by the lack of empirical research. Objectives: To employ the flicker paradigm to investigate social information processing bias among sexual offenders with a learning disability, non-offenders with a learning disability and non-offenders without a learning disability. Design: An independent groups design was employed to compare the data obtained from a participant group of sexual offenders with a mild learning disability with two comparison groups of non-offenders with a learning disability and non-offenders without a learning disability. Method: Participants were asked to view successively and repeatedly on a monitor two versions of a visual scent (an original and a slightly altered version of the original) until they detected the change. The changes to the original visual were either sex-related (midriff of a person being exposed) or neutral related (object being removed from a scene) changes. The experiment was counterbalanced with participants viewing equal number of sex-related and neutral related changes to the original stimuli. Results: Non-offenders without a learning disability required less time to identify sex-related and neutral related changes than individuals with a learning disability. However, no significant interaction was observed between type of participant group and type of stimulus change. This finding indicated that the nature of change in the stimulus did not influence the speed at which the participants responded. Results are discussed in relation to previous research. Conclusion: Results obtained from the present study highlight that this is a complex area to investigate, as there is currently no clear framework to guide this area of research. Strengths and limitations of the present study are addressed and areas of future research are explored.