Abstract: In order to promote the development efficiency of tobacco flavors, a multi-objective optimization genetic algorithm based on optimal solutions of non-dominated sorting genetic algorithm (NSGA-Ⅱ) was used to establish a digital tobacco flavoring method with the help of descriptive information database of 344 monomer fragrance materials and 30 empirical formulas of tobacco flavors. Taking floral and fruity scents as two flavoring targets, flavor formulas were successfully formed by the established digital flavoring method. The results showed that:1) The obtained floral and fruity flavor were consistent with the corresponding target requirements both in olfactory aroma and cigarette flavoring. 2) The olfactory aroma of the floral flavor was mainly floral note (7.0 points) and sweet note (5.0 points). Adding floral flavor remarkably promoted the scores of floral, sweet and fresh scents of cigarette smoke from 0.1, 2.5 and 2.0 points to 2.5, 3.3 and 3.0 points, respectively. 3) The olfactory aroma of the fruity flavor was mainly fruity note (7.0 points) and sweet note (5.5 points). Adding fruity flavor significantly increased the scores of fruity, sweet and fresh scents of cigarette smoke from 0.5, 2.5 and 2.0 points to 3.2, 3.5 and 3.1 points, respectively. 4) Cigarette flavoring experiments verified that the two flavors created by the digital method were harmonious with the tobacco blends, improving the sensory quality of cigarette smoke and meeting the requirements for tobacco flavors. 5) This digital flavoring method simplified the flavoring process and thus promoted the efficiency of cigarette flavoring.