In addition to mental retardation (MR), fragile X [fra(X)] has been associated with other developmental disabilities, autism in particular. Recently, several studies have concluded that the association of fra(X) with autism is at best weak and perhaps nonexistent. This study examined reports of previously published data from an epidemiological perspective to determine if the prevalence of fra(X) among autistic males was significantly different from that among MR males. Nineteen studies of autistic males and 21 studies of MR males were analyzed from 59 examined. Of 5601 MR males tested, 307 (5.5%) were cytogenetically positive. Of 1006 autistic males, 54 (5.4%) were positive. Analysis indicated no statistical difference between these proportions (Z = 0.014; p greater than .50). An odds ratio (OR) was also computed to determine the risk of autism from fra(X). Estimated OR approximately 1.0 [0.73, 1.32]. Attributable risk (AR), the proportionate excess risk of autism associated with exposure to fra(X) was AR approximately 0.0. That is, there was no attributable risk of autism from fra(X). The association of fra(X) with autism may reflect the association with MR that generally accompanies autism, as several earlier reports have noted. However, the similarity in prevalence of fra(X) among autistic and MR males may be biased in the studies which find no instances of fra(X) among autistic males represent 12% of the pooled sample.