Aiming at the problem that the supply of emergency material distribution depots might be disrupted after a disaster, this paper studied the post-disaster emergency material supply considering the disruption risk of distribution depots. In pursuit of globally optimal solutions, decisions including location selection of distribution depots, the assignments of material demands to distribution depots, and the supply sequencing at each distribution depot were integrated into a combinatorial scheduling-location problem. A scenario-based mixed-integer linear programming model was formulated with the objective function of minimizing the expected makespan of the supply task. A matheuristic algorithm based on predetermined ordering and kernel search (MPOKS) was proposed to solve the problem efficiently. A numerical experiment is conducted and the result shows that the MPOKS proposed in this paper is significantly effective and can provide precise, quantitative, and globally optimal scheduling plans for a task with up to 200 disaster-affected areas and five distribution depots within reasonable computation time.