The legume genus Glycine is an economically important lineage due to the inclusion of soybean (G. max), one of the world’s most widely grown crops.  There are approximately 30 perennial species in subgenus Glycine with diverse habits. The group remains taxonomically challenging due to informally recognized taxa, as well as previous classification of both diploids and tetraploids as G. tomentella s.l. All Glycine species share two rounds of ancient polyploidy, one ~65 Ma shared with all papilionoid species and a unique event occurring ~10 Ma. In addition to the ancient polyploidy events there have been at least eight different recent allopolyploid (< 1 MY) events involving diploid perennial species. Using a combination of GBS and genome resequencing data from three different polyploid complexes (G. pescadrensis, G. tabacina, and G. tomentella) we investigate the evolutionary history of recent allopolyploidy in the genus. Using sequencing data we can accurately discriminate suspected allotetraploid accessions from diploid accessions based on genome size and heterozygosity values. However, due to the repeated rounds of polyploidy that Glycine has undergone and the level of gene duplication present in reference genome assemblies, estimating ploidy from sequencing data alone is impossible. First, using the coding regions from the available diploid reference genomes, we show varying degrees of incongruence of single copy loci and many instances of gene duplications and loss. By creating in silico tetraploid genomes, we were able to call SNPs to investigate phylogenetic relationships of sampled tetraploid accessions, as well as the repeated formation of tetraploids. By dating both chloroplast and nuclear based phylogenies, we obtain a better idea of when the duplication events occurred and provide hypotheses for the incongruence of some taxa. We show the utility of combining different data sets to address phylogenomic questions in taxa with nested polyploidy, while also highlighting potential pitfalls.