The issue of sexual offending in people with intellectual disabilities (ID) is a poorly researched field, and the existing models do not take into consideration the particular cognitive and social factors that determine the offending behavior. The Counterfeit Deviance model suggested in the present research will attempt to address this gap by providing a conceptual framework that incorporates thinking errors, social perceptions, and coping mechanisms to explain sexual offending among people with ID. This paper will test the applicability and empirical validation of the model. A mixed-method design has been used. Forensic psychologists, caregivers, and individuals with ID whose qualitative interviews (N=30) were developed to form model components. The correlations among model constructs were investigated based on a quantitative survey (N=200) by regression analysis and structural equation modelling (SEM). Sexual offence risk was the dependent variable and cognitive, emotional, and social predictors were the independent variables, which were tested using path analysis to test predictive validity of the model. The Counterfeit Deviance model has shown a high predictive ability (ΔR² = 0.48, p< 0.01), with cognitive distortions (β = 0.35, p< 0.05) and social perceptions (β = 0.42, p< 0.01) having a considerable influence on sexual offending behavior. The model accounted a significant amount of the variance of the risk of offending compared to the traditional models (ΔR²= 0.12, p = 0.05). The Counterfeit Deviance model provides a new and strong model of viewing sexual offending in individuals with ID. The relevance of its use in theory and practice is provided by its empirical validation leaving the basis of future research. This research is an improvement of clinical assessment and intervention of sexual offenders with ID, which shapes treatment strategies that consider certain cognitive and social variables. Further studies are needed to elaborate on the model by investigating how it can be used in various settings and populations, as well as to determine its applicability in the outcome of treatment.