Advances in genotyping and related technologies and the study of families with two or more autistic siblings have allowed the reporting of various genetic linkage and candidate gene association studies for autism. Despite this significant research effort, there still is no general agreement on the role of particular genes that predispose to the risk of autism. One source of general debate is whether core symptoms studied in research on the heredity of autism include single or multiple independent genetic domains. In this review, it is argued that failure to control for co-existing dimensions of psychopathology and lack of assessment of parents may lead to misinterpretation of genetic findings from pedigrees involving multiple affected family members.