Grey optimum analysis and grey equal weight cluster analysis were used to compare the chemical compositions and quality of flue-cured tobacco leaf samples from southwest China. Results showed that the sugar–protein ratio, sugar–nicotine ratio, total amount of ether extract, total alkaloid content, pH, protein content, and potassium–chloride ratio were the chemical composition indices most closely related to the sensory quality. These were used to evaluate and compare the tobacco leaf quality by grey equal weight cluster analysis. Based on the grey equal weight cluster analysis, the 30 samples of flue-cure tobacco from southwest China were divided into two grey categories. The grey categories were verified by using them to selecting similar and different substitutes for a target tobacco in a mixed tobacco formula. The new formula with the similar substitute produced comparable sensory quality results to the original formula, and formula containing the different substitute had dissimilar sensory qualities to the original formula. These results confirm that the joint application of grey optimum analysis and grey equal weight clustering has a high degree of confidence for comparison of tobacco leaf quality.