Existing research on optimizing electric bus charging station locations often assumes an exogenous demand, overlooking the feedback effects of station locations on demand. Moreover, the long-term implications of location strategies are deeply influenced by the complex interactions between land-use and transportation systems. To address these two challenges simultaneously, this study develops a bi-level programming model—a hierarchical decision-making framework involving two interconnected problems. Specifically, the upper-level problem is formulated as a mixed integer nonlinear programming model that minimizes the electric bus system's investment, operation, and passenger waiting time costs by optimizing the fleet size of electric buses, the corresponding frequency setting, and the location and capacity of charging stations. The lower-level model is an integrated land-use and transportation model that captures the long-term impacts of upper-level location decisions on transportation and land-use systems. To solve the proposed model, an iterative solution method is devised, which employs Gurobi to generate upper-level decisions via solving a linearized upper-level model and subsequently evaluates the decisions via TRNUS, which is an integrated land-use and transportation model, in the lower-level. Case studies are carried out using real data from Jiangyin City, China. The results demonstrate that the optimal design considering the interaction between land use and transportation attracts a higher number of bus users across various routes and increases the share of passenger kilometers traveled by bus from 19.9 % to 20.5 %. Meanwhile, it contributes to alleviating traffic congestion by 2.7 %, improving regional accessibility by 0.4 %, and reducing vehicle carbon emissions by 1.1 %, promoting urban sustainability.