Designs for improved railroad track structure are needed to ensure safety, reliability and profitability as the railroads strive to compete with other means of transport. The large number of precast sleepers produced in the country calls for an optimal design inthe interest of economy. Use of optimisation methods can provide a scientific approach toarrive at a leastcost design. In this paper, optimal design of broad gauge prestressed concrete sleepers for tangent tracks is presented considering the current design standards by Research Designs and Standards Organisation (RDSO). Formulations for optimal design for the conventional 22.9 t and proposed 25 t axle loads are presented along with identification of design variables, objective function and constraints. Optimisation was carriedout in two phases: first, by keeping the geometry constant (for a constant mould size, as presently used) and secondly, also by varying the sleeper cross-section geometry. Among the non-traditional optimisation techniques, genetic algorithm for a constrained optimisation problem using adaptive penalty parameter and adaptive genetic algorithm was used. A MATLAB program has been developed based on the formulations. The optimal solutions are summarised here.