Zhao et al. [2007] in their "Unified Theory of Autism" hypothesized that incidence of autism in males could be explained by essentially two types of family structures: majority of autism cases are from low-risk autism families with de novo mutations, and a minority of cases are from high-risk multiplex families, where risk to male offspring approximates 50% consistent with a dominant model and high penetrance. Using the Autism Genetic Resource Exchange (AGRE) data set, Zhao et al. identified 86 high-risk families with likely dominant transmission. As genotype data are now available for many members of the AGRE resource, the objective of this manuscript was to determine if dominant linkage evidence for an autism predisposition gene exists in these 86 high-risk families. HumanHap550K Illumina SNP data were available for 92% of 455 total family members in these 86 high-risk families. We performed a linkage analysis using a pruned subset of markers where markers in high linkage disequilibrium were removed. We observed a single suggestive peak (maximum LOD 2.01, maximum HLOD 2.08) under a dominant model on chromosome Xp22.11-p21.2 that encompasses the IL1RAPL1 gene. Mutations or deletions in IL1RAPL1 have been previously reported in three families with autism. In our study, 11 families contributed nominally (P<0.05, HLOD>0.588) to the chromosome X peak. These results demonstrate that identification of a more homogeneous subset of autism cases, which was based on family structure in this study, may help to identify, localize and further our understanding of autism predisposition genes.