Conduct problems in adolescence is a well-known phenomenon where a range of negative behaviors can have multiple detrimental effects on the youth themselves, their family members and surroundings. Callous-unemotional (CU) traits have been linked to more severe and persistent antisocial behaviors and are characterized by a lack of empathy, a lack of remorse and guilt, a lack of concern in performance and shallow affect (Frick et al., 2014). Research suggests that there might be a distinct developmental pathway for conduct problems in children with CU traits, evidenced by both neurobiological and -psychological indicators (Frick et al., 2014). This has warranted the addition of the CU-specifier “with limited prosocial emotions” to the conduct disorder. CU traits might thereby provide a more nuanced assessment of risk factors associated with conduct problems and could assist in providing more optimal treatments for youth with these characteristics.This PhD-dissertation had three aims related to CU traits in adolescents. The first aim was to assess the construct validity of a commonly used measure of CU traits, the Inventory of CallousUnemotional traits (ICU), in a Norwegian sample of at-risk adolescents. The second aim was to assess the degree to which CU traits are linked to treatment outcomes among adolescents with behavior problems who receive Functional Family Therapy (FFT). The third aim was to assess the degree of change in CU traits for adolescents participating in FFT. The analyses were based on data from a single group pre-post evaluation of FFT in Denmark (N=407), and a randomized controlled trial of FFT in Norway (N=159-160). Aim 1: The construct validity of the Norwegian translation of the ICU. Previous research is somewhat inconsistent regarding the factorial structure of the ICU and the items on the Unemotional subscale have been particularly scrutinized. The first paper included in this thesis (Paper 1) thus studied the construct validity of the Norwegian translation of the ICU. A multi-informant sample of 160 adolescents and their parents and teachers provided the data for this study. The results from Paper 1 showed that a two-factor solution of a Callousness factor and an Uncaring factor provided the best fit to the model across respondent groups. The Callousness factor was most strongly related to anti-social behavior, while the Uncaring factor was most strongly related to the lack of fear. Despite these distinct correlational patterns, the two-factor structure could be the result of a common method-variance effects related to whether items were standard- or reverse-scored. It is therefore likely that CU traits are best operationalized as a unidimensional measure. Aim 2: The relationship between CU traits and treatment outcomes. The degree to which CU traits impact treatment outcomes for adolescents with behavior problems was the focus of the remaining papers of this thesis (Paper 2 and Paper 3). In both of these there was little evidence to suggest that CU traits were linked to diminished treatment outcomes. On the contrary, CU traits were linked to steeper decreases in anti-social behavior and steeper increases in prosocial behavior in Paper 2. In Paper 3, the relationships between CU traits and treatment outcomes became non-significant when controlling for baseline levels of behavior problems. One possible explanation for these findings is that adolescents with CU traits have more problems at baseline and thereby a greater potential range of improvement. Importantly, the results in Paper 3 showed that although CU traits were either neutrally or positively related to treatment outcomes, baseline CU traits were still predictive of problematic behavior at the end of treatment. This suggests that adolescents with elevated levels of CU traits could be in need of more tailored and long-term services to sustain behavioral improvements over time. The impact of CU traits on treatment outcomes was not different when comparing FFT treatment to an active control condition, when only looking at those with clinical levels of externalizing problems at the start of treatment or when comparing primary to secondary CU subtypologies.Aim 3: The malleability of CU traits. With respect to whether CU traits change over the course of treatment, the results from Paper 2 and Paper 3 were somewhat mixed. In both papers the overall change in CU traits across treatment was small, and a reliable decline in CU traits was observed in relatively few adolescents. At the same time, both papers found more pronounced reductions in CU with respect to both magnitude and the proportion of participants with reliable decreases in CU traits among the subset of adolescents with increased CU traits at baseline. This suggests that CU traits could act as a moderator of the malleability of CU traits. There were no differences in the malleability of CU traits when comparing FFT treatment to an active control condition and when comparing primary to secondary CU sub-typologies.Conclusion: The results from this PhD-dissertation point to the need for continued development of measures of CU traits in both research and clinical settings. Furthermore, the results suggest that adolescents with behavior problems and CU traits do benefit from evidence-based and high-quality treatment, through observed declines in aggressive and rule breaking behavior as well as increases in social skills. Such behavioral improvements seem obtainable without a reduction in the CU traits themselves, as the results indicate that CU traits seem to be relatively stable across time even in the context of treatment. Future work should study the possible benefits of prolonged interventions that specifically target CU-associated risk factors in adolescents, to see if behavioral improvements beyond those obtainable in evidence-based programs and high-quality treatments can be reached. The way in which a reduction in CU traits can be achieved and lead to long-term positive outcomes is also of great importance.