In this paper we present an allocation model for a particular container supply chain where multiple sea-terminal quays and an inland terminal are considered. From the moment a container is unloaded from the vessel at the sea-terminal, several options for the nal transportation to the hinterland can be considered: trucks, trains and barges. The logistic planner decides which is the best allocation for the containers to the available eet. The objective is to minimize the transportation costs while guaranteeing a high service level. In general, the planner tries to bundle as many containers as possible and load them on barges or trains. Nevertheless, time constraints, such as due dates and dierent arrival times, make the bundling tricky and force the allocation to faster, more exible and more expensive modes of transport as trucks. We build a model for this setting, which has its roots in the classical Heterogeneous Fleet Vehicle Routing Problem (HVRP). We propose a case-study of an inland terminal in the Netherlands, whose data-sets are processed and solved by means of a Markov chain Monte Carlo heuristic. We explain in detail the characteristics of the dierent decisions and we provide insights for the real-world application.