This study was to investigate near-infrared reflectance spectroscopy (NIRS) application for determination of individual soluble carbohydrate content in soybean seeds for development of high-quality new soybean variety and quality control of soy products. NIRS was used as a rapid and simultaneous analysis method to determine individual soluble carbohydrate content of soybean seeds in Korea. A total of 240 samples (about 2 g) were scanned in the reflectance mode and the reference values for the individual soluble carbohydrate contents were measured by high performance liquid chromatography. Calibration equation for individual soluble carbohydrate contents were developed using modified partial least-squares regression with internal cross validation (n=155). The equation for sucrose content had lower standard error of cross-validation, higher $R^2$ (coefficient of determination in calibration), and higher ratio of unexplained variance divided by variance (1-VR) value than that for raffinose, and stachyose. Prediction of an external validation set (n=85) showed significant correlation between reference value and NIRS estimated value based on the standard error of prediction (SEP), $r^2$ (coefficient of determination in prediction), and the ratio of standard deviation (SD) of reference data to SEP. The model for sucrose had higher value of RSP (ratio of standard deviation for prediction, the ratio of SD of reference data to SEP(C) in the external validation set) and $r^2$ (3.50 and 0.921, respectively), thereby that equation has good quantitative accuracy, whereas those of raffinose, and stachyose had lower value (below 0.89 and 0.45, respectively). These results indicated that NIRS could be used to rapidly determine the content of sucrose in soybean seed samples in the breeding programs for development of high-quality new soybean variety.