Based on genetic algorithm, an improved method for scheduling arrival aircrafts at an airport with multiple runways was studied. Scheduling arrival aircrafts is a typical non linear combinatorial optimization problem, and the application of genetic algorithm to this problem was explored.A sequence of aircrafts was a chromosome. We designed the crossover operator and mutation operator. In decoding phase, first we assigned the runway, then we used the improved moving method to remove the conflict between aircrafts, insert idle time, decide the sequence and time of aircrafts landing. The simulation result demonstrated that the presented algorithm can assist controllers to make decisions.