Designing large logistics systems, such as automated container terminals, is a challenging task that requires the collaboration of multiple organizations. To support the design process, simulation models are used to provide insight into the behavior and performance of future systems. Using design science research, we propose a multi-actor simulation-based design method for logistics systems, which we call the Multiple Worlds method. The design process follows a tree structure: each node contains a design that can be developed further in a next node depending on specific design decisions. Each design can be directly linked to appropriate simulation components from different simulation component libraries. Using different component libraries, designers are guided through the design process and stimulated to generate alternative designs. To instantiate and evaluate the method, we have developed a software environment that consists of different sets of simulation components developed using ESDEVS D-SOL, an application to support the construction of the Multiple Worlds tree, and a 3D visualization application to visualize the results. We have evaluated this method and results indicate that the method can indeed be used to collaboratively design logistics systems and can be useful in business environments.