Increased living standards and high occupants comfort demands lead to a growing air-conditioning market. This results in increased energy consumption and an increased CO2 emission. Thus in the EU, 40% of the total energy consumption and 20% of the total CO2 emissions are due to building sector. Besides, the CFCs used in the air conditioning devices have negative impact on environment and on occupants health. For these reasons open desiccant cooling cycles have been developed as an alternative to vapour compression cycles. The main purpose of this study is to evaluate the potential of cooling and the performance of a standard desiccant cooling air-handling unit coupled with a solar installation. This unit comprises a desiccant wheel in tandem with a thermal wheel with evaporative coolers in both air supply and return air streams before the thermal wheel. This system allows cooling and dehumidifying air without using conventional refrigerants. The desiccant wheel contains a desiccant material (Lithium Chloride) which needs to be regenerated with an external heat source. This heat is taken from a solar installation consisting of a solar storage tank and solar collectors. Since the required regeneration temperatures are low (40°C to 70°C) it is possible to use flat plate collectors. To evaluate system performance and its cooling potential, it should be coupled with a building for different climatic conditions and different building configurations. This evaluation requires an object oriented program. That is why models are developed with SPARK, an equation based modelling environment. The main advantage of this environment is that models of individual components are input/output free. In this article, we study the functioning of the system coupled with a building located in Chambery (east of France). First, we present the system; describe its components and their modelling in SPARK environment. Then, we describe experimental system which allowed validating modelling of the desiccant wheel. Finally we use simulations to evaluate system cooling potential. A graphic aided method called “Boundary lines” method was used and its results were also validated using seasonal simulations.