Water distribution system (WDS) design is considered as a class of large combinatorial non-linear optimization problems having complex implicit constraints such as conservation of mass and energy equations. Due to the complexity and large feasible solution, traditional optimization techniques are not capable to tackle these kinds of problems. Recently, applications of metaheuristic algorithms, due to their efficiencies and performances, are increased dramatically. In this paper, water cycle algorithm (WCA), a recently developed population-based algorithm, coupled with hydraulic simulator, EPANET, are applied for finding the optimal cost design of WDS. The performance of the WCA is shown using well-known Balerma benchmark problem widely used in the literature. The obtained optimization results using the WCA are compared with other optimizers such as genetic algorithm, simulated annealing, and harmony search. Comparisons of obtained statistical results show the superiority of the WCA over other optimization methods in terms of convergence rate and solution quality.