Performance on a single executive function (EF) task (e.g., a card sorting task) is often taken to represent ability on the underlying subcomponent of EF (e.g., set shifting) without accounting for the non-specific and non-executive skills employed to complete the task. This study used a manualised battery of EF tasks to derive individual task scores and latent EF scores. Seventy-nine adolescents aged between 11 and 19 years, including 37 autistic and 42 non-autistic participants, matched on cognitive ability, completed the battery. Autistic adolescents had moderate global EF difficulties and had significantly more difficulties on some individual tasks. However, the samples did not differ on any of the specific individual subcomponents of EF (fluency, cognitive control and working memory).