Background: Individuals at ultra-high risk (UHR) for psychosis exhibit deficits in social functioning that may relate to efficacy of social skills performance. There is, though, a scarcity of objective measures assessing social skills performance in the UHR state. This study assessed the psychometric properties of a relatively new social skills measure, the High Risk Social Challenge Task (HiSoC), in an UHR population. Methods: Exploratory factor analysis was used to assess the factor structure of the HiSoC task in 102 UHR individuals and 66 healthy controls (HC). Convergent and discriminant validity of the HiSoC was assessed. Results: Our findings revealed a three-factor structure comprising the factors “Affect”, “Odd Behaviour and Language”, and “Social-Interpersonal”. The test showed excellent inter-rater reliability (ICC between 0.88 and 0.98). The HiSoC correlated significantly, as expected, with measures of global and social functioning demonstrating construct validity and utility as a social skills assessment tool in the UHR population. The HiSoC task discriminated between UHR and HC with large effect sizes (Cohen's d range = 1.40–1.94). Discussion: Our findings provide evidence for a three-factor structure of the HiSoC task corresponding to the original American version of the task. Additionally, our findings reveal the HiSoC to be sensitive to social skills impairments in the UHR population and suggest it to be a suitable screening tool for social skills deficits in UHR states. The robust correlations with real-life functioning indicate social skills to be an important target for assessment and intervention within the UHR population.