The alexithymia hypothesis posits that specific aspects of the autistic socio-cognitive profile, namely emotion recognition difficulties, can be explained by the increased prevalence of alexithymia in autistic populations. However, this hypothesis has largely been tested in adults. We tested whether co-occurring alexithymia could account for associations between autistic traits and two key domains of socio-emotional cognitive functioning in adolescence. Participants were 184 (44% male) children aged 10–16 years with and without a diagnosis of autism (n = 75 autistic), who completed experimental measures of emotion recognition and theory of mind. Parents completed assessment of autistic traits and alexithymia. Pre-registered analyses tested associations between autistic traits and socio-cognitive task performance, and whether including alexithymia as an additional predictor changed the nature of the association between autistic traits and task performance. Results showed participants with higher levels of autistic traits performed lower on emotion recognition and theory of mind tasks, and that these associations were not significantly altered by including alexithymia in the models. Results suggest that more work is needed to understand the applicability of the alexithymia hypothesis in younger populations, but that at least in young adolescents and when using parent-report measures, alexithymia may not be associated with emotion recognition or theory of mind ability. Lay abstract Alexithymia is a sub-clinical condition characterised by difficulties identifying and describing one’s own emotions, which is found in many, but not all autistic people. The alexithymia hypothesis suggests that certain aspects of socio-cognitive functioning typically attributed to autism, namely difficulties in emotion recognition, might be better explained by often co-occurring alexithymia. It is important to understand what is specific to autism and what is due to other co-occurring characteristics to develop appropriate support for autistic people. However, most research on this topic has been conducted in adults, which limits our knowledge about the relevance of this theory to younger autistic populations. This study tested whether difficulties in emotion recognition and theory of mind traditionally associated with autism might be better explained by alexithymia in a sample of adolescents with and without a diagnosis of autism. Results found that difficulties in emotion recognition and theory of mind were both associated with autistic traits, and this was not accounted for by individual differences in levels of alexithymia. This research suggests that more work is needed to understand the applicability of the alexithymia hypothesis in younger populations, but that at least in adolescents and when using parent-report measures, alexithymia may not account for emotion recognition or theory of mind difficulties associated with autistic traits.