High rates of alexithymia, a condition characterised by difficulties identifying and describing emotions, are frequently reported in both children and adults with autism spectrum disorder (ASD). However, the dilemma of measuring alexithymia via self-report has rarely been addressed. In this study, we compared objective and subjective measures of alexithymia in adults with ASD and typically developing adults. We found significantly higher levels of alexithymia in the ASD sample as measured by both self-report on the Toronto Alexithymia Scale (TAS-20) and by the observer rated Alexithymia Provoked Response Scale (APRQ). However, the two measures did not correlate with each other. We explore the different facets of the alexithymia construct that these two measures may be distinguishing.