The potential for greater energy efficiency gave rise to the popularity of implementing district cooling systems. In newer districts, however, the discrepancy between the designed capacity of the cooling system and actual cooling demand usually negates these benefits. In such scenarios, the optimization of the system's operations with respect to cooling demand could considerably improve the energy efficiency of the system, without incurring additional capital costs. Components of a district cooling system are usually operated at pre-defined setpoints or individually optimized, without regard of the impact on the overall system. Formulation of an optimization problem which adequately captures the thermal and physical interactions as well as the tight coupling between components, i.e., holistically, results in a mixed integer non-linear program which is large and difficult to solve. In this article, a hierarchical optimization framework for the hourly operation of district cooling systems is introduced to manage the problem. The initially complex model of the system was abstracted so that it could be solved effectively using the combination of a genetic algorithm and mixed integer linear program. The mixed integer linear program reduced the search space of the genetic algorithm, thereby increasing the likelihood of achieving global optimality. Finally, the methodology was applied to a case study based on an existing district cooling system in Europe for illustrative purposes. For the scenarios defined, the thermal and physical variables for each component were tuned such that the hourly cooling demand could be fulfilled with minimal electricity consumed. Results indicate potential electricity savings of up to 31%. At the optimum, some components operated less efficiently for the benefit of the overall system, further reinforcing the advantage of performing optimization holistically.