In order to find out the correlation between tobacco quality evaluation indicators in China's traditional "moderate aroma" tobacco-producing areas and simplify the tobacco quality evaluation indicators, we evaluate the appearance quality and smoking quality of 143 flue-cured tobacco leaf samples in China's "moderate aroma" tobacco-producing areas, test the physical traits and chemical component, and analyze the canonical correlation between four quality evaluation indicators. The results show that there is significant or extremely significant canonical correlation between four evaluation indicators (tobacco smoking quality, chemical component, appearance quality and physical trait quality); the cumulative variance contribution rate of evaluation indicators is in the order of chemical component (69.17%)>appearance quality (68.76%)>physical traits (64.13%); appearance quality is most closely related to physical traits (93.84%). The individual indicators for tobacco quality evaluation make different contribution to the correlation between quality evaluation indicators. The chemical component evaluation indicators mainly include total sugar and ratio of total sugar to betaine; sensory taste indicators mainly include aroma volume, smoke concentration, irritation and softness degree; physical trait evaluation indicators mainly include leaf weight, leaf length and leaf density; appearance quality indicators mainly include leaf organizational structure, color, maturity and identity. Studies have shown that in the large-scale eco-region, using canonical correlation analysis to simplify tobacco quality evaluation indicators is feasible.