This thesis reports on improvements in controlling solar driven adsorption chillers. Since simulation is used as main research method, a part deals with the modelling of the main system components. The control problematic is investigated in a first approach considering the chiller alone and in a second approach on the system level. The potential for performance improvement of the chiller through the control of internal parameters is investigated. The adsorption/desorption duration has a strong influence on both cooling capacity and efficiency of the chiller. Different ways of performing heat recovery have been investigated. The concept of mass recovery is also a good way to increase the chiller efficiency. Then, the improvement potential of controlling the adsorption/desorption duration during chiller operation is investigated. The characteristic equations of sorption chillers have been modified in order to account for varying operating conditions over a wide range of inlet temperatures and adsorption/desorption durations. Finally, a method to control solar driven adsorption chillers by adjusting dynamically several control variables is proposed. The first control goal with the highest priority is to adjust the chiller cooling capacity according to the building cooling load. The second goal is to minimize the system power consumption and is achieved by the use of an optimization algorithm. Based on the developed method, several control strategies that act on different control variables (hot and cooling water temperatures, adsorption/desorption duration) are investigated through simulations of a solar cooling system. The implementation of these advanced controls is simulated first on daily basis and then the best strategies are investigated over the whole cooling season.