A new methodology and computational model are developed for direct evaluation of an optimal storm water inlet design. The optimal design is defined as the least-cost combination of inlet types, sizes and locations that effectively drain a length of pavement. Costs associated with inlets are user-defined and can include those associated with materials, installation and maintenance, as well as other project-related expenses. Effective drainage is defined here as maintaining a spatial distribution of roadway spread, or top width of gutter flow, that is less than the allowable width of spread. The solution methodology is based on the coupling of a genetic algorithm and a hydraulic simulation model. The simulation model follows design guidelines established by the Federal Highway Administration and is used to implicitly solve governing hydraulic constraints that yield gutter discharge, inlet interception capacity and spread according to a design storm. The genetic algorithm is used to select the best combination of design parameters and thus solves the overall optimization problem. Capabilities of the model are successfully demonstrated through application to a hypothetical, yet realistic, highway drainage system. The example reveals that genetic algorithms and the optimal control methodology comprise a comprehensive decision-making mechanism that can be used for cost-effective design of storm water inlets and may lead to a reduced overall cost for highway drainage.