Flood control detention facilities have been traditionally designed using the design storm approach. Because of the deficiencies associated with the design storm approach, continuous simulation using long-term historical rainfall data has been recommended for the planning and design of these facilities in order to examine the performance of the system under a wide range of meteorological and hydrological conditions. An analytical probabilistic approach is presented as a computationally efficient alternative to continuous simulation. In this approach the probability distribution of the peak outflow rate from a detention facility servicing an urban catchment is derived from the probability distribution of the rainfall event characteristics that generate the runoff received by the detention facility. These derived mathematical expressions are used to determine analytically the storage-discharge relationship required for a detention facility to achieve the desired level of flood control. Comparison with long-term continuous simulation modeling demonstrates that the closed-form mathematical expressions of the analytical probabilistic approach provide reasonable approximation of the continuous simulation results. The analytical probabilistic approach overcomes some of the conceptual problems of the design storm approach and is therefore proposed for the planning and design of flood control detention facilities. ¿ 1999 American Geophysical Union