A study was carried out to accurately and rapidly identify planting areas of flue-cured tobacco. A total of 201 flue-cured tobacco samples from three different areas in Kunming, Honghe and Qujing, Yunnan Province were selected for the study. After collecting the near-infrared spectra of different areas and reducing the interference factors through the spectral preprocessing method, followed by principal component analysis (PCA) for dimensionality reduction, a whale algorithm (WOA) was established to optimize support vector machine (SVM) parameters to establish an automatic identification method. In the wavenumber range of 8000 to 4000 cm^-1, the standard normal variable transformation (SNV) combined with the second derivative method (2D) was used for near-infrared spectroscopy preprocessing, and the data after the PCA dimensionality reduction was used as the input variable, after which the WOA-optimized support vector parameters could achieve a better recognition effect. The classification accuracy rate of the training set is 97.18 %, and the classification accuracy rate of the test set is 98.31 %. This shows that using near-infrared spectroscopy technology combined with WOA algorithm to optimize SVM can achieve accurate identification of area of flue-cured tobacco.