Résumé : Agent-Oriented Simulations (AOS) are helpful in the understanding of real complex systems particularly thanks to the ease of modeling and observation they provide. To fully exploit those simulations, AOS platforms should offer shorter simulation times in spite of the in- creasing complexity of simulation models. In this article, we propose a method to mini- mize execution time of AOS. This method is based on the identification of agents’ temporal dependencies. Knowing these dependencies, we use a network of AOS platforms to provide parallel execution of simulation agents. Our proposition is based on the Temporality model we previously presented to ensure events causality during simulation.