Due to the requirements of the production process and storage of Chinese herbal extracts, a certain amount of maltodextrin is added. At present, there is no standard method for determining the content of maltodextrin in Chinese herbal extracts in China and overseas. This study aimed to establish a method for rapid determination of the maltodextrin content in the mixed Pueraria extract powder. Using 88 samples of Pueraria lobata extract powders containing different contents of maltodextrin as the exptl. materials, and the mid-IR spectra of samples with known maltodextrin contents were collected. Baseline correction and normalization pretreatment were performed on the IR spectra, and the quant. model was established based on the partial least squares regression method. The results showed that the quant. model had good prediction effect, and the prediction model had strong stability and high prediction precision with the decision coefficient (R2) as 0.9819 and standard deviation as 3.205. The mid-IR spectroscopy technique can quickly quantify the content of maltodextrin in Pueraria lobata extract, and effectively monitor the extract market to avoid substituting shoddy products for quality products, thereby providing tech. support to the market supervision department.