The STK39 gene encodes a serine/threonine kinase of 547-amino acids ubiquitously expressed in the brain and pancreas, and it may function in the cellular stress response pathway during states of hypotonic stress. An earlier report provided evidence for an association between autism and the single nucleotide polymorphism (SNP) rs1807984 in the STK39 gene (Ramoz et al. 2008). This led us to investigate, using the same sample, whether any of the core autism domains (i.e. impaired social interaction and communication, and restricted and repetitive behavior) or, more specifically, any of the four autism-related features previously identified as familial (Silverman et al. 2001) (i.e., circumscribed interests, routines and rituals, level of language, and age at phrase speech) varied by genotype for this SNP (Table 1). With parental written informed consent, we included all affected autism spectrum disorder (ASD; including autism, borderline autism (Silverman et al. 2001), Asperger’s disorder, pervasive developmental disorder—not otherwise specified) cases ascertained at our center who had previously been genotyped for the rs1807984 allele (Ramoz et al. 2008). As described elsewhere (Silverman et al. 2001), we used the Autism Diagnostic Interview—Revised (ADI-R) (Lord et al. 1994) to classify 336 children (mean age = 8 ± 5 years) derived from 181 families (multiply affected families: n = 141) with either autism (n = 289) or another ASD (n = 47). Each family had at least one child who met criteria for ICD-10/DSM-IV autism. The ADI-R also allowed us to quantify each participant’s symptoms. The primary set of models included all ASD cases to examine the familiality of the autism symptom clusters in an analysis of covariance (ANCOVA) using SPSS version 17, where family membership was treated as a random effect and age and sex were covariates. The ADI Domain Scores and the four previously identified familial autismrelated features did not differ by genotype (all p values [ 0.15). A secondary set of ANCOVA models focused solely on the 289 subjects from these same families who met full criteria for ICD-10/DSM-IV autism (211 cases of autism were from 103 multiply affected families). Again, we saw no differences in domain scores or the additional four autism features based on the risk allele genotype (all p values [ 0.30). Finally, a tertiary set of ANCOVA models focused again on pure autism cases but used only one, randomly selected, affected family member (n = 181) as an alternative method to avoid dependent observations due to family clustering. Here too, no significant differences were observed (all p values [ 0.20). Collectively, our results suggest no significant association between any of the core autism symptom domains or the four additional previously identified familial features and the rs1807984 SNP on the STK39 gene. It is important to note that the inability to observe a specific relationship with these autism traits among ASD cases does not contradict the previously observed evidence R. D. Vavolizza University of North Carolina, Chapel Hill, NC, USA