Abstract: Gasoline blending model is a key factor for ensuring the qualified rate of gasoline product, as well as to improve profit. In this paper the quantitative relationship between physicochemical property and hydrocarbon group composition of gasoline is figured out to study the optimization of gasoline blending model, and the concrete forms of objective functions and constraint conditions are provided. According to the characteristics of this model, it is solved by adjoint equation method. Case studies of adopting this model show that the precision of model prediction is high, thus, it can provide technical foundation for optimizing gasoline blending and predicting product quality indexes.