In China,aroma types of flue-cured tobacco were divided into light,medium,and heavy.However,evaluation of flue-cured tobacco aroma types using chemical compositions in a quantitative way at national scale is limited.This study investigated the spatial distribution of flue-cured tobacco aroma types across China using back propagation (BP) neural network and multi-year average chemical compositions of the flue-cured tobacco leaves.The overall accuracy produced by BP neural network was 80.11%,and the kappa coefficient was 0.70.These indicated that the classification model produced substantial agreement.The results and methodologies are expected to provide valuable information on regional planning and decision making for production of high-quality flue-cured tobacco.