The use of near-infrared (NIR) spectroscopic sensing system for the prediction of milk quality indicators in cow milk was described in this study. The objective of this study was to investigate the measurement accuracy of the sensing system developed in our study. Three major milk constituents (fat, protein, and lactose), milk urea nitrogen (MUN) and somatic cell count (SCC) of two Holstein cows belonging to Hokkaido University dairy farm were measured. Milk spectra with a wavelength range of 700 to 1050 nm and milk samples were collected in every 20 seconds during milking using the NIR spectroscopic sensing system. Calibration models were developed using partial least squares (PLS) regression analysis and the precision and accuracy of the models were validated. The statistical results obtained for milk fat and protein contents were very high while the results obtained for milk lactose, MUN and SCC were sufficiently high. This suggest that the NIR spectroscopic sensing system developed in this study could be used for online real-time milk quality determination of each cows’ milk constituents, MUN and SCC during milking. This sensing system could assist dairy farmers in solving the challenge of effective individual cow management, resulting in smart dairy farming.