Abstract: Objective: To establish a method for the consistency evaluation and quantitative analysis of near infrared spectroscopy(NIRS) of ginger formula granules. Methods: NIR spectral information of ginger formula granules samples were collected using a portable NIR spectrometer. The fingerprints of 61 batches of ginger formula granules samples were tested for consistency with the control fingerprints, and evaluated using correlation coefficients. The content of 6-gingerol in the ginger formula granules samples was determined by high performance liquid chromatograph(HPLC). The effects of different preprocessing methods on the parameters of the quantitative model were investigated, and the partial least squares(PLS) and support vector regression(SVR)quantitative models were established. Results: In the consistency evaluation, the correlation coefficients between the fingerprints of the 61 batches of ginger formula granules and the control fingerprints were all greater than 0.97 in the consistency test, the batch-to-batch consistency of the ginger formula granules samples was good.Among the established quantitative models, the optimal PLS model had a root mean square error of correction(RMSEC) of 0.0711, along with a correlation coefficient of correction(R_C) of 0.6240. Additionally, the root mean square error of prediction(RMSEP) for this model was 0.0654, and the correlation coefficient of prediction (R_P) was 0.7009. For the SVR model, the RMSEC was 0.0448 with an R_C of 0.9198, while the RMSEP was 0.0663,and the R_P stood at 0.7765. Conclusion: The SVR quantitative model has better predictive performance than the PLS quantitative model, and can rapidly predict the content of 6-gingerol. This research performs effective, rapid and nondestructive quality analysis of ginger formulation granules based on the portable near infrared spectrometer, which provides a new method reference for the quality control of ginger formula granules.