Because of the importance of multimodal transport for goods importation and exportation from Morocco to international customers, our goals are modeling and solving multimodal freight transportation problems by more focusing on some issues confronted by our country. After presenting the required mathematical background on robust and multiobjective optimization and surveying the literature works on multimodal transport problems, its variants and solution approaches; and instead of testing them only on traditional benchmarks; we dealt with new variants and methods. A multi-objective mathematical model is proposed to optimize the multimodal transport of agricultural products from Morocco to Europe. To solve this model adapted to a real network, two multiobjective version algorithms are presented in order to optimize simultaneously the total transportation cost and the over-time.Given the complexity of the global logistics chain by various operations and several stakeholders, the data of the problem may sometimes be ignored or uncompleted. That’s why we have developed a mathematical model for multimodal terminal location problem by integrating uncertainties of the transportation costs, the capacities and the using costs of terminals. A population based simulated annealing and a hybrid method combining an exact and a metaheuristic approaches are proposed to effectively solving the problem. The test is then performed on a real intermodal network. These contributions are concluded by remarks and highlights about the impact of perishability and robustness on flows, location strategies of ports, the choice of paths and transportation modes.