<p>This study proposes an intermodal freight transportation network design problem (IFTNDP) with specific features. Essentially, IFTNDP is a leader–follower game between a network planner (e.g., a government), as the leader, and users (e.g., carriers or freight forwarders), as the follower. In this study, a set of short-term and long-term decisions including road pricing, road fuel pricing, and hub location are made by the network planner to maximize container-kilometer (TEU-km) on rail links connecting intermodal hubs. Users, as the follower, react to the network planner’s decisions and choose appropriate routes for their freight containers. This study proposes a probability-based route choice model (i.e., a binary logit model) developed based on a state preference method, which is integrated into the IFTNDP by a set of constraints. Therefore, the model can visualize users’ preferences for various route factors such as cost, time, reliability, or safety. Furthermore, the model adopts a multi-user platform rather than a single-user platform to enhance the level of disaggregation and subsequently the quality of results. According to the high complexity of the proposed model, a Lagrangian relaxation and sub-gradient algorithm (LRSA) is applied to efficiently treat large-scale problems. The model is tested on a real network and six practical observations made by various sensitivity analyses are presented. Hence, as an example, the results demonstrate that a 20% increase in the costs of truck paths by road pricing can significantly contribute to increasing the TEU-km on the rail links and reducing the number of necessary intermodal hubs in the network.</p>