With 17 Gb and an hexaploid composition, the bread wheat (Triticum aestivum L.) genome remains largely unknown and its molecular characterization remains a challenge. Our current knowledge on the composition and organisation has been limited to the analysis of small DNA fragments (150 kb) and of random BAC end sequences that have revealed an impressive richness in transposable elements, ca 65-80%. To improve our knowledge on the structure of the wheat genome and the impact of the repetitive sequences on its structural and functional evolution, we have sequenced a dozen of Mb-sized contigs (1-3 Mb) originating from different regions of chromosome 3B for which a physical map has been recently established in the laboratory. One of these contigs belongs to a multiple fungal disease resistance region and encompasses the leaf rust resistance locus rph7 that corresponds to a highly rearranged locus in grasses (1, 2). It is composed of 26 BACs for which sequencing and assembly are nearly completed and that are currently annotated. With 50 predicted genes (1 gene/60 kb in average), the rph7 locus appears as a gene-rich region. Known TEs represents 58% of the sequence, the majority of them being LTR-retrotransposons (72%) and CACTA family transposons (20%). In addition, the curated annotation of repeated sequences has led to detect new elements. The genes and TEs are not randomly distributed along the locus: the length of intergenic regions varies from 500 bps to 230 kb. A global synteny can be observed with the rice genome for the entire contig. However, among the 50 predicted genes, only 20 are conserved at orthologous positions in rice whereas the 30 other genes share similarity with genes located at different loci in the rice genome. Detailed comparative analyses are now in progress for a better understanding of the rearrangement events and of the molecular mechanisms that have shaped this highly variable locus as well as the role of TEs in its variability.