A stepwise regression method was introduced to study the relationship between cigarette sensory flavor and quality and organic acids in 25 cut tobacco samples.To identify organic acids key to sensory flavor and quality index,variable selection was performed for 15 measured organic acids.The regression models were applied in three trial productions of cigarette samples,and the predicted sensory flavor and quality scores were close to the true values,indicating that the selected organic acids variables in the models were well associated with the corresponding sensory flavors and quality indexes.All the results showed that organic acids in cut tobacco,such as isopentanoic acid,malonic acid,phenylacetic acid and citric acid,had either positive or negative effects on different sensory flavors and quality indexes.