Oakley, Brewer, Bird, and Catmur (2016) investigated whether the Reading the Minds in the Eyes Test (RMET) measures emotion recognition rather than theory of mind (ToM). To explore this, 19 participants with autism and 23 controls, matched on alexithymia traits, were tested with the RMET, as well as the ToM Movie for Assessment of Social Cognition (MASC). The authors found a significant difference between the two groups on the MASC but not on the RMET, but dividing the groups based on alexithymia resulted in a significantly lower performance on the RMET but not on the MASC for the alexithymia group. Therefore, they conclude that difficulties on the RMET are associated with alexithymia, not autism, while difficulties on the MASC are associated with autism, not alexithymia. Here we investigated what seems to be opposite patterns of performance on the two cognitive tasks within the autism group, which modified the authors' interpretation of their data. This was examined by correlating the alexithymia scores with the RMET and a subscale of the MASC scores, referred to as the cognitive MASC. We found a negative correlation between the alexithymia score and the RMET score while also finding a positive correlation between the alexithymia score and the cognitive MASC score in the autism group. Such an opposite pattern of performance suggests the presence of distinct patterns of ToM difficulties within the autism group. This also indicates that, contrary to what is reported by Oakley et al., there is an association between alexithymia and the MASC within the autism group. (PsycINFO Database Record (c) 2019 APA, all rights reserved).