Soybean () is one of the main sources supplying human dietary protein, and increasing its protein content is one of major targets in soybean breeding. Identification of the genes modulating the protein content is of great significance in breeding of soybean varieties with higher protein content. In this study, the F2:16 and F2:17 recombinant inbred lines (RIL) populations derived from ‘Zhonghuang 35’ and ‘Tokachi nagaha’ were applied to delimit the previously-mapped protein content locus qPRO-19-1. By sequencing the candidate genes in the interval, an InDel marker in Glyma.19g223300 between parents and five polymorphic markers from 22 SSR markers were developed. The physical interval harboring qPRO-19-1 was further delimited from 384 kb to 68.03 kb, including four annotated candidate genes, of which Glyma.19g221800 and Glyma.19g222000 showed extremely significant difference on transcript level at different stages of seed development. These results provided a basis for future map-based cloning and molecular marker-assisted breeding of this protein content-related gene in soybean.