The nondestructive method and system for predicting fruit sugar content with near infrared diffuse reflectance technique were explored. The influence of different testing distances on measuring accuracy of sugar content (SC) of Fuji apples was discussed. Diffuse reflectance spectra were measured in the spectral range from 12500 to 4000 cm^-1 at four different testing distances of 0 mm, 2 mm, 4 mm and 6 mm. Principal component regression (PCR) analysis was used to develop the calibration models between the original spectral data and measured SC. The results showed that the best distance for measuring sugar content was about at 0 mm, and the calibration model was set up at this distance. The calibration model gave the correlation coefficient of 0.844, with a standard error of calibration of 0.729, a standard error of prediction of 0.864, and a bias of 0.318. The study shows that the near infrared diffuse reflectance technique was feasible to nondestructively measure the sugar content of apple.