Technical Abstract: Restriction Fragment Length Polymorphism (RFLP) mapping data from nine separate populations (G. max X G. soja and G. max X G. max) of the Glycine subgenus soja genome has led to the identification of many duplicated segments of the genome. Linkage groups contained up to 33 markers that were duplicated on other linkage groups. The size of homoeologous regions spanned by three or more duplicated markers ranged from 8.7 cM to 107.6 cM, with an average size of 45.7 cM. One would expect to observe two sets of homoeologous sequences in a diploidized tetraploid, which the soybean has been assumed to be. However, we observed segments in the soybean genome that were present in as many as six copies with an average of 2.55 duplications per segment. Quantitative trait loci for seed protein and oil showed correspondence across homoeologous regions, suggesting that the genes or gene families contributing to these characters have retained similar functions throughout the evolution of the chromosomes. Many examples of duplicated segments nested within other duplicated segments were observed. These results suggest that at least one of the original genomes contributing to the polyploid genome of the glycine subgenus soja may have undergone an additional round of tetraploidization, and/or that large internal duplications in addition to tetraploidy are responsible for the highly duplicated nature of the genome of the subgenus. When the identification of all duplicated sequences is complete, saturation mapping of those segments should allow the transfer of map information from well-mapped to poorly-mapped regions of the genome.