Multi-Objective Particle Swarm Optimization (MOPSO) algorithm based on the concept of Pareto optimality was used to efficiently evaluate the thermal performance of a conventional adsorption ventilated cooling system. A desiccant wheel numerical model was developed and then coupled to a thermodynamic one in order to predict the different states of airflow through the system components. Instead of the dehumidification performances of the desiccant wheel, the present study focused on optimizing simultaneously the COP and the cooling capacity of the system. The process and regeneration air velocity, the rotational speed of the desiccant wheel and the regeneration temperature were considered as the main initial working parameters to generate the particle swarm within the research space. The results showed that the developed MOPSO algorithm is suitable for this kind of complex optimization problem and allowed, with a minimum computational effort, to generate the Pareto front solutions considered as the best compromises between the objective functions. The use of a purge zone during the optimization process improves the cycle performance mainly by lowering the regeneration heat and the pre-cooling of the process air. The application of MOPSO seems to be useful for the cooling system operators, providing them with a wide choice of operating conditions that will meet the predefined cooling requirements with a better COP.