Linkage disequilibrium (LD) is the "non-random association of alleles" and can be utilized through association analysis to discover quantitative trait loci (QTLs).  If a population has extensive LD, few markers will be needed to scan the whole genome for QTLs, but the positions of these QTLs will not be well defined.  Conversely, in a population with limited LD, genetic factors can be fine mapped.  Our goal was to provide an initial assessment of LD in four distinct soybean populations: Glycine soja, the wild soybean; Asian G. max; North American cultivar ancestors; and North American public cultivars released in the 1980s.  Multiple fragments throughout three 300+kb regions were sequenced for common SNPs (freq. >0.10) in the four soybean populations.  The three regions were located on the soybean linkage groups A2, G, and J surrounding the rhg1, Rhg4, and Rps2 disease resistant loci, respectively.  The extent and structure of LD throughout the three genomic regions were similar in the Glycine soja population, with average LD extending approximately 60 kilobases (kb).  The other three populations exhibited different patterns of LD ranging from 100 kb to >600 kb depending on the genomic region and the soybean population being tested.  The variability of LD structure and extent around these three genomic regions complicates the prospects of applying whole genome genetic association analysis in soybean without first assaying the genome for LD structure.  Such an analysis will assist in determining which regions require only low marker density and which regions need high marker density for a thorough genome scan.  Wild soybean appears to have enough degraded LD to allow for “fine mapping” (within a 60 kb region) of QTLs or candidate genes identified in cultivated soybean with its more extensive LD.