This study selected total acid content and pH value as characterization indicators and used hyperspectral imaging technology and chemometric methods to discriminate uniformity of Zhenjiang balsamic vinegar grains. First, hyperspectral transmission images in 432~960nm of vinegar grains, total acid content and pH value were acquired. Secondly, PLS and LS-SVM method were used to establish uniformity indicators content prediction models after preferred variables which were selected by GA and siPLS. The root mean square error of prediction (RMSEP) and correlation coefficient (R) for the test set are 0.389% and 0.7751 for total acid content, 0.0417 and 0.7974 for pH value. Total acid content and pH value in each pixel point on the vinegar grains were obtained by the above prediction models. After pseudo-color processing, the distribution map of total acid content content and pH value before and after turning up the vinegar grains were obtained. In the distribution maps, the range of total acid content and pH value in the grain were 3.0% to 7.8%, 3.5 to 4.2 before the overturning, and 4.8%~7.0%, 3.6~3.9 after the overturning. By overturning the grains, the degree of uniformity was improved while high local concentrations phenomenon still exists. The overall results revealed that hyperspectral imaging technology is a promising technique to discriminate the degree of uniformity of grains after overturning rapidly and nondestructively.