Callous-unemotional (CU) traits identify adolescents at high risk for severe and recurrent antisocial behaviour and are included in the DSM-5 as a specifier to conduct disorder. The Inventory of Callous-Unemotional Traits (ICU) might be the most cost-effective screening instrument for detecting CU traits in youth. We aimed to address an important gap in research by comparing the factor structure and psychometric properties of ICU scores across gender, age, and setting (community versus institutionalized). Community-based (n=397) and institutionalized (n=164) adolescent boys and girls completed self-reported measures of psychopathic traits (including the ICU), externalizing problems and aggression, and a laboratory measure of emotional recognition; the staff reported antisocial behavior for institutionalized children. Factor-analytic results indicated that a three-factor bifactor model best fit the data across samples, with measurement invariance across gender, age, and setting, supporting the construct validity of the ICU. In support of its criterion validity, across groups ICU scores were positively correlated with analogous dimensions from the Youth Psychopathic Traits Inventory, measures of proactive aggression, and self-and staff-reported antisocial behavior, and were associated with poorer accuracy in recognizing distress-related emotions. Our findings thus support the overall utility of the ICU for assessing CU traits in adolescent populations regardless of gender, age, and community or institutionalized status, but suggest a need for refining its assessment of the shallow affect dimension.