Abstract: Assuming that the inventory costs in cold chain increase linearly and time-dependently with mass loss and that the selling prices of refrigerated goods rise slowly and come up to the peak at a certain time, then the prices are going down slowly, an optimal model is developed for storage/sales of refrigerated goods, where the objective is maximizing the profits after the stocks have all been sold out with the selling time taken as a decision variable. The conditional extreme value method of nonlinear programming is used to solve the model. Simulation results verified the feasibility and effectiveness of the model by a numerical example.