--- - |2+ The aim of the present study was to assess whether the nature of the main autistic features (i.e., social communication problems and repetitive and restrictive patterns) are better conceptualized as dimensional or categorical in a school population. The study was based on the teacher ratings of two different age groups: 2,585 children between the ages of 10 and 12 (Primary Education; PE) and 2,502 children between the ages of 3 and 5 (Nursery Education; NE) from 60 mainstream schools. The analyses were based on Factor Mixture Analysis, a novel approach that combines dimensional and categorical features and prevents spurious latent classes from appearing. The results provided evidence of the dimensionality of autism spectrum symptoms in a school age population. The distribution of the symptoms was strongly and positively skewed but continuous; and the prevalence of high‐risk symptoms for autism spectrum disorders (ASD) and social‐pragmatic communication disorder (SCD) was 7.55% of NE children and 8.74% in PE. A categorical separation between SCD and ASD was not supported by our sample. In view of the results, it is necessary to establish clear cut points for detecting and diagnosing autism and to develop specific and reliable tools capable of assessing symptom severity and functional consequences in children with ASD. Autism Res 2018, 11: 979–988. © 2018 International Society for Autism Research, Wiley Periodicals, Inc. Lay Summary The results of the present study suggest that the distribution of autism spectrum symptoms are continuous and dimensional among school‐aged children and thus support the need to establish clear cut‐off points for detecting and diagnosing autism. In our sample, the prevalence of high‐risk symptoms for autism spectrum disorders and social‐pragmatic communication disorder was around 8%. - Autism Research, Volume 11, Issue 7, Page 979-988, July 2018.