Hexaploid wheat ( Triticum aestivum L em Thell) is derived from a complex hybridization procedure involving three diploid species carrying the A, B and D genomes, respectively. We recently isolated microsatellites from a T. tauschii library enriched for various motifs and evaluated the transferability of these markers to several diploid species carrying the A, B or D genomes. All of the primer pairs amplifying more than one locus on bread wheat and half of those giving D-genome-specific loci gave an amplification product on A-and/or B-diploid species. All of the markers giving a single amplification product for T. tauschii and no amplification on the other diploid species were D-genome-specific at the hexaploid level. The non-specific microsatellite markers (which gave an amplification product on diploid species carrying the A, B or D genome) gave either a complex amplification pattern on bread wheat (with several bands) or generated a single band which mapped to the D genome. Southern blot hybridizations with probes corresponding to the microsatellite flanking regions gave a signal on all diploid and hexaploid species, whatever the specificity of the microsatellite. The patterns observed on bread wheat were generally in accordance with those observed for diploid species, with slight rearrangements. This suggests that the specificity of microsatellite markers is probably due to mutations in microsatellite flanking regions rather than sequence elimination during polyploidization events and that genome stringency is higher at the polyploid than at the diploid level.