The prevalence of fragile X syndrome (FXS) is approximately 7% in Thai boys with developmental delay of unknown cause. To determine if FXS might have a specific haplotype association, we analyzed 125 unrelated control subjects and 25 unrelated FXS patients using 3 microsatellites, DXS548, FRAXAC1 and FRAXE, and two single nucleotide polymorphisms, ATL1 and IVS10. FRAXAC1 and DXS548 are located approximately 7 kb and approximately 150 kb proximal to the CGG-FMR1 whereas ATL1, IVS10 and FRAXE are located approximately 5.6 kb, approximately 24.5 kb and approximately 600 kb distal to the CGG-FMR1. We found 40 haplotypes in the control group and 14 haplotypes in the FXS group. Of 14 haplotypes in the FXS group, 6 haplotypes were not found in the control group suggesting possible new mutations or admixture of immigrant haplotypes. We observed that most diverse haplotypes came from different FRAXE alleles. For this reason, we analyzed haplotypes composed from the remaining markers alone (DXS548-FRAXAC1-ATL1-IVS10). We found 2 major haplotypes (20-18-G-T and 20-19-A-C) with no significant haplotype differences between the control group (67/125 of 20-18-G-T and 25/125 of 20-19-A-C) and FXS group (16/25 of 20-18-G-T and 6/25 of 20-19-A-C). The other haplotypes found were 33/125 in the control group and 3/25 in the FXS group. The two major haplotypes associated FXS in Thai subjects were the two most common haplotypes in the normal Thai subjects. We could not prove, therefore, that there were founder effects at the FRAXA locus in Thailand. We could not, however, exclude it completely. These findings apparently contrast with most other reports on FXS founder effects in various ethnic groups.