To improve signal timing and traffic flow operation at the intersection with contraflow left-turn lane, the application condition for the contraflow left-turn lane was developed. The arrival and discharge pattern of left-turn vehicles were divided into 8 groups and the calculation method for left-turn vehicle delay under each group was carried out. The average intersection vehicle delay minimum was taken as the objective, cycle length, green times of main and pre-signal phases, and the lengths of contraflow left-turn lanes were taken as optimization variables, and then the signal control plan optimization model was developed. The genetic algorithm was used to solve this model. A case study was conducted to validate the optimization method, which was also compared with the traditional control method. Results show that the optimization method can increase the capacity of left-turn vehicles, decrease the required green time and then reduce the intersection cycle length. The decrease in the green time of left-turn phases would result in the decrease in the red time of through phases and then the decrease in intersection average delay. In peak hour, average vehicle delay can be decreased by 23.8%.