In order to improve the coordinated observation efficiency of space and air observation resources, this study proposes a two-stage iterative optimization method to solve the coordinated task planning problem of space and air resources, which is under the divide and conquer framework. In the first stage, the fitness function is formed according to two indices, ie., observation opportunity and conflict degree. And tasks are assigned to different sub-planners based on the fitness vaule. In the second stage, each sub-planner schedules observation resources according to the assigned tasks and generates corresponding task planning schemes. The resources observation schemes and related observation profit are returned to the first stage. Iteratively, the resources observation scheme is adjusted and updated to generate approximate optimal or satisfied task assignment schemes and resources observation schemes by using a tabu-list based optimization strategy, with the consideration of the feedback of task planning shemes. In order to verify the effectiveness of the proposed two-stage iterative optimization method, this study has done extensive simulated experiments. The results show that the proposed method can increase the total task profits by about 2.40%sim14.14%, compared with the highest weight first assignment(HWFA) algorithm, the task assignment algorithm based on fitness value(TAAFV) and the task assignment algorithm based on resource priority(TAARP). The achievement of this study can realize the effective coodinated task planning for space and air observation resources.