It is very important to examine the effect of non-genetically modified (non-GM) soy varieties, which constitute the matrix of the testing sample used to quantify GM soy (RRS), on the measured value of RRS by quantitative PCR methods. Therefore, we quantified the amount of RRS in powder-mixed samples containing 1 or 5% RRS prepared by using 10 different varieties of non-GM soy as the matrix. The results revealed that the measured values were not in agreement with the powder-mixing levels and that the extent of the difference depended on the variety of non-GM soy used as the matrix. The yields of DNA extracted differed among the soy varieties. On the other hand, analysis of DNA-mixed samples, that were prepared with the DNAs extracted from RRS and non-GM soy varieties, showed that the measured values of RRS were in agreement with the DNA-mixing levels. These results strongly suggest that the proportions of DNA derived from RRS and non-GM soy were not consistent with the powder-mixing ratio in the case of some non-GM soy varieties used as a matrix, resulting in the discrepancy between the measured values and the powder-mixing levels.