Pedigree analysis provides information helpful in the preservation of genetic diversity in soybean [Glycine max (L.) Merr.]. An often-used, but untested assumption in pedigree analysis of soybean is that the pedigree relationship between ancestors of unknown origin is zero. We investigated genotypic relationships among 14 ancestors that constitute = 70% of the genetic base for North American public soybean cultivars. The experiment was conducted at the Southeastern Plant Environmental Laboratories phytotron under controlled light and temperature conditions. The ancestral strains of soybean exhibited a wide range of genetic diversity for 10 metric traits; more than that predicted by a previous study employing molecular markers. Multivariate analysis collapsed the 10 traits into four principal components that explained 80% of the total genotypic variation among the strains. Similarity estimates (s) for all pairs of the ancestral types were derived from the four components. Subsequently, coefficient of parentage (r) estimates (derived from pedigree analysis) were calculated for 258 modern soybean cultivars. A first pedigree analysis employed the standard assumption given above and a second analysis substituted s values in place of zero as estimates of genetic relatedness in the founding stock. The similarity estimates did not have a large effect on estimates of r for cultivars released after 1980. Thus, the assumption of zero relation among founding stock should not influence soybean breeders' approaches to the use or preservation of genetic diversity. The small impact of s on the estimation of r is theorized to result from the infrequent occurrence of extremely high or low s values between the nine ancestors that are major contributors to modern soybean. A small r was observed between the modern southern and northern cultivars, indicating that these two gene pools remain distinct and offer the best available reservoirs of quantitative gentic diversity for practical breeding.