Aronia melanocarpa is a small berry listed in the list of new food raw materials,rich in anthocyanins and other ingredients,which has been widely used in alcohol,beverages,functional food,cosmetics and other fields,with high economic value. Due to the influence of environmental factors such as climate and planting conditions in different areas,the fruit quality of A. melanocarpa is significantly different. Therefore, to standardize the market management of A. melanocarpa fruit,the fruit of A. melanocarpa from different places of origin was identified by mid-infrared spectroscopy combined with chemometrics. 750 infrared spectral data of A. melanocarpa fruit from 15 production areas were collected. After spectral pretreatments,such as multiple scattering corrections (MSC),standard normalization (SNV),moving smoothing (SG),first derivative (FD),second derivative (SD),and so on,the optimal spectral pretreatment method was determined by comparing the recognition effect of support vector machine (SVM) modeling with the original spectrum. The K-S sample division method divides the samples into training sets and test sets at a ratio of 4∶1,and then the samples are normalized. The competitive adaptive reweighting algorithm (CARS) and continuous projection algorithm (SPA) are used to extract the spectral feature information,and the best model is determined by modeling and comparing with random forest (RF),extreme learning machine (ELM) and support vector machine (SVM). The results show that MSC is the best spectral preprocessing method;the recognition rate of the MSC-SVM training set is 93.33%,and the recognition rate of the test set is 92.67%,which can effectively reduce the random error generated during spectral acquisition. After extracting the MSC characteristic spectral wavenumber by CARS and SPA,the modeling results of the three algorithms are compared,and the SPA-SVM model is determined to be the best recognition model. The recognition rate of its training set and test version is 100%,and only 16 wavelength points are needed to complete the accurate recognition. Therefore,the combination of mid-infrared spectroscopy and chemometrics,especially SPA-SVM model,can accurately identify the origin of A. melanocarpa fruit,provide a fast and simple method support for the origin traceability and quality evaluation of A. melanocarpa fruit,and provide a technical basis for building a unique brand with regional characteristics.