In this study, a portable, non-destructive tester, integrating near-infrared (NIR) spectroscopy, was developed to determine the maturity of Asian pears. In addition, regression analysis was conducted for mature Manpungbae pears using the partial least squares regression model with the Brix value of Asian pears in a wavelength range of 710-930 nm, and the standard error of calibration or standard error of prediction (SEC/SEP) of 0.41-0.43 degrees Brix and relative standard error of calibration or relative standard error of prediction (rSEC/rSEP) = 0.91-0.92 were observed, indicating excellent prediction performance with no dependence on pretreatment conditions. Furthermore, the reliability of the developed Brix determination model was confirmed using Manpungbae, Wonhwang, Chuhwangbae, and Niitaka pears, and SEP, rSEP, and bias values of 0.29-0.33 degrees Brix, 0.90-0.96, and 1.67E-04, respectively, were observed, indicating high reliability. Hence, the construction of a Brix determination model by the accumulation of absorbance for several varieties of pears in each harvest period is expected to determine optimum harvest times via the measurement of the Brix values at full maturity regardless, of the variety. (C) 2017 Elsevier B.V. All rights reserved.