The work aims to study the loading problem of vehicle logistics with the minimum number of car carrier as the goal so as to reduce logistics cost of the car. Mathematical model of the mixed loading of the car was established firstly. Different mixed loading schemes were compared and analyzed by enumeration algorithm. Schemes with loading rate more than 90% were screened out to establish a combination matrix. Then, the mathematical model of the minimum number of car carrier was set up by using the car's full load scheme and the cars'number and cars'type as the constraint conditions. Finally, the genetic algorithm was used to solve the optimal soluton, and the calculation results were verified. The simulation results showed that the optimal loading scheme calculated by genetic algorithm was close to the optimal solution obtained by enumeration, but the computing time of the former was only about 1/200 of the latter. In conclusion, the genetic algorithm has fast convergence and closer results with the theoretical optimal solution. It has advantages of both computing time and calculation effect.