In order to solve the problem of incomplete disaster information caused by information lag in the early phase of the earthquake, which leads to the lack of consideration for disaster urgency in aviation rescue dispatch schemes, a method of aircraft rescue dispatch that combines seismic disaster risk was proposed with the consideration of disaster urgency. A model of disaster urgency based on seismic risk was constructed and combined with geographic information system (GIS) to evaluate the differential urgency of affected areas using the analytic hierarchy process (AHP) and entropy weight method, thereby establishing time window constraints. Therefore, a penalty function was introduced to construct an optimization model for aircraft rescue scheduling with the objectives of shortest rescue time and minimum rescue cost. Finally, Sichuan was designed as the background of the arithmetic example, and the non-dominated sorting genetic algorithm-Ⅱ (NSGA-Ⅱ) was employed to conduct numerical simulations and comparative analyses for multi-objective optimisation problems. The results show that the model constructed is able to better predict the urgency of each disaster area in the early stage of the earthquake, ensure the priority and focus of the rescue force. Compared with traditional scheduling schemes, the method proposed in this paper only slightly increases rescue time while significantly reducing penalty costs and total costs, and minimizing secondary losses caused by delays. It offers distinct advantages in controlling rescue costs and mitigating losses from delays, and can provide important theoretical foundations and practical guidance for the optimal allocation of aviation emergency rescue resources during emergencies such as earthquakes.