Abstract  Virtual Worlds are facing several challenges to become a mature technology commonly used in their numerous potential applications such as games, training, education, health and collaborative work. Main challenges are the introduction of multimedia in virtual worlds and the decrease of the cost and the required time for their design, which remains an empirical process. This paper proposes a methodology for introducing multimedia and decreasing the cost and needed time. The main interest of the proposed methodology is to provide virtual worlds with multimedia, interoperability and reusability properties. Interoperability enables virtual entities belonging to virtual worlds independently designed (i.e., no prior explicit agreement exists between theirs respective designers) to interact with each others. Reusability enables virtual worlds designers to use a virtual entity pattern initially designed for a virtual world A to be used in the design of a new virtual world B. The originality of this methodology relies on multiagent concepts and learning techniques for avoiding prior complex specifications tasks to achieve interoperability and reusability of multimedia virtual entities in virtual worlds.