The heterogeneity of symptoms among patients with autism spectrum disorder (ASD) is a major problem for researchers working on the etiology and pathology of the disorder. This diversity also poses a significant hurdle in evaluating the efficacy of potential therapies. Various attempts at grouping ASD patients based on behavioral symptoms have been made. However, Tony Persico and colleagues in Rome have used a much larger set of criteria, based on a questionnaire encompassing 36 developmental, clinical and family history variables. In a new paper, they describe a hierarchical clustering test of their ability to place patients into one of four categories: (i) prominent immune abnormalities accompanied by some circadian and sensory issues (18% of the patients fell into this category), (ii) major circadian and sensory dysfunction without immune symptoms (18% of patients), (iii) stereotyped/repetitive behaviors and abnormal social behaviors (31% of patients), and (iv) mixture of all four components with great disruptive behaviors and mental retardation (34% of patients). The most robust category turned out to be the immune dysregulation group, which is characterized by a much greater history of allergies, immune disease or food allergies in the patient and/or their relatives, pre-term birth rate, history of obstetric complications and/or recurrent abortions in the mother, lower birth weight, and greater rate of infectious disease at time of autism diagnosis. Head circumference and height is greater in this group compared to patients in the other groups. It is also striking that GI problems are prominent in this group compared to the other groups. In light of our mouse model, it would be of great interest to determine if maternal infection was more frequent in this immune dysregulation group.