AbstractBackground:DSM-5 proposed a new operational system by using the number of fulfilled criteria as an indicator of gambling disorder severity. This method has proven to be controversial among researchers and clinicians alike, due to the lack of studies indicating whether severity, as measured by these criteria, is clinically relevant in terms of treatment outcome. Additionally, numerous studies have highlighted the associations between gambling disorder and impulsivity, though few have examined the impact of impulsivity on long-term treatment outcomes.Methods:In this study, we aimed to assess the predictive value of DSM-5 severity levels on response to cognitive-behavioral therapy (CBT) in a sample of male adults seeking treatment for gambling disorder (n = 398). Furthermore, we explored longitudinal predictors of CBT treatment response at a follow-up, considering UPPS-P impulsivity traits.Results:Our study failed to identify differences in treatment outcomes between patients categorized by DSM-5 severity levels. Higher baseline scores in negative urgency predicted relapse during CBT treatment, and higher levels of sensation seeking were predictive of drop-out from short-term treatment, as well as of drop-out at 24-months.Conclusions:These noteworthy findings raise questions regarding the clinical utility of DSM-5 severity categories and lend support to the implementation of dimensional approaches for gambling disorder.