Abstract: As the key element of automatic test systems, the Test Task Scheduling Problem (TTSP) is becoming an increasingly important issue. The TTSP often involves two kinds of constraints. The temporal constraints among tasks are often the net ones and the resource constraints are generated due to the lack of enough resources for each task. The current methods to handle the constraints can be classified into optimization and satisfaction. Most of them suffer much difficulty to solve the constraints in the TTSP. However, the optimization methods are abandoned since the infeasible solutions are not allowed in the TTSP to avoid the accidents. The current satisfaction methods are usually for the tree or chine constraints. In addition, there is seldom research on the temporal and resource constraints together. Hence, the Priority-based (PB) encoding is provided to solve the temporal constraints while the Scheme Choice Rule (SCR) is created to find the corresponding schemes to distribute certain resource to these tasks. The compounding of PB and SCR called PBSCR is embedded into a genetic algorithm and applied to two benchmarks. Empirical results suggest that PBSCR outperforms better than not only the optimization methods but also the satisfaction methods. The efficacy of PBSCR is thereby justified.