Common variants contribute significantly to the genetics of autism spectrum disorder (ASD), although the identification of individual risk polymorphisms remains still elusive due to their small effect sizes and limited sample sizes available for association studies. During the last decade several genome‐wide association studies (GWAS) have enabled the detection of a few plausible risk variants. The three main studies are family‐based and pointed at SEMA5A (rs10513025), MACROD2 (rs4141463) and MSNP1 (rs4307059). In our study we attempted to replicate these GWAS hits using a case‐control association study in five European populations of ASD patients and gender‐matched controls, all Caucasians. Results showed no association of individual variants with ASD in any of the population groups considered or in the combined European sample. We performed a meta‐analysis study across five European populations for rs10513025 (1,904 ASD cases and 2,674 controls), seven European populations for rs4141463 (2,855 ASD cases and 36,177 controls) and five European populations for rs4307059 (2,347 ASD cases and 2,764 controls). The results showed an odds ratio (OR) of 1.05 (95% CI = 0.84–1.32) for rs10513025, 1.0002 (95% CI = 0.93–1.08) for rs4141463 and 1.01 (95% CI = 0.92–1.1) for rs4307059, with no significant P‐values (rs10513025, P = 0.73; rs4141463, P = 0.95; rs4307059, P = 0.9). No association was found when we considered either only high functioning autism (HFA), genders separately or only multiplex families. Ongoing GWAS projects with larger ASD cohorts will contribute to clarify the role of common variation in the disorder and will likely identify risk variants of modest effect not detected previously. Autism Res 2017, 10: 202–211. © 2016 International Society for Autism Research, Wiley Periodicals, Inc.