A two-stages successive genetic algorithm was used to optimize the scheduling of TT&C (Tracking Telemetry and Command) resource from space and land. Because the object was somewhat separable, the scheduled time windows were separated into two sections. After the population from first section was evolved, the gained optimized solution was combined with the second section, and then the evolution of second phase goes further. For lessening the local searching limitation, virtual time windows were set with some probability in the fitness computation process of first phase, which could retain the individuals that might be the component of overall optimization solution. Simulation demonstrates the proposed method can get good solution at the cost of less time.