Background and aims: Previous research investigated the prevalence and risk factors of problematic sexual behaviour (PSB) using the Bergen–Yale Sex Addiction Scale (BYSAS), among other instruments. However, a dearth of literature employed item response theory (IRT) to assess the BYSAS psychometric properties. The present study adopts an IRT framework to comprehensively examine the measurement aspects (including discrimination and severity) and the prevalence of PSB among a relatively large adult sample. Methods: Participants (N = 968, 64.3% men, Mage = 29.5 years, age range = 18–64 years) completed the BYSAS. Results: CFA determined that the BYSAS is a unidimensional construct. Additionally, IRT analysis showed variability in discrimination, severity, and reliability across BYSAS items, with a raw score exceeding 20, indicating a high risk of PSB. Accordingly, using this raw score 1.8% of the participants were at-risk of PSB. Conclusions: Findings supported the differential use of BYSAS criteria for assessment purposes, while only a minority of participants presented to be at risk of problematic sexual behaviour difficulties.