Previous studies of founder chromosome effects in fragile X have been based on linkage disequilibrium with either FRAXAC1 or DXS548 alone or combined with FRAXAC2. Recently, we found no linkage disequilibrium of FMR-1 with FRAXAC2, but rather, found FRAXAC2 was complex and highly mutable. Therefore, we have now analyzed FRAXAC1 and DXS548 together for haplotypes, two markers which have not been jointly analyzed previously, to test for disequilibrium. We typed 315 fragile X (FX) chromosomes and controls, further subdivided into large controls (LC) and small controls (SC) with < or = 35 repeats and identified 26 different haplotypes. Two were more frequent and one less frequent in FX than SCs, thus confirming apparent linkage disequilibrium in fragile X. However, we noted increased FX microsatellite heterozygosity, either individually (results quite similar to previous studies) or as haplotypes. This heterozygosity covaried with FX > LC > SC, which may indicate alternative explanation exists for the apparent disequilibrium. We hypothesize that large FMR-1 CGG repeat allele genes may be associated with the generation of new microsatellite mutations. Possible mechanisms include gene conversions between CGG repeats and flanking microsatellites involving unequal double cross-overs, the expansion of small control CGGs to larger sizes associated with episodic generalized microsatellite instability or as a direct result of mutant FMR-1 gene function. We conclude that the founder effects observed with the use of these CA repeats is likely to reflect both linkage disequilibrium and increased microsatellite instability of fragile X chromosomes.