Designing optimal water supply systems is an important purpose of any urban system that involves relevant installation, operation and maintenance costs. However, achieving the optimal design is known to be a complex task, indeed the corresponding mathematical model for this problem leads to a non-linear and non-convex problem classified as NP-hard. In this paper, we propose using the cuckoo search algorithm which a modern bio-inspired metaheuristic based on the obligate brood parasitic behavior of cuckoo birds. This behavior is combined with the interesting Lévy flight, which mimic the exploration of some birds and flies, that move by combining straight flights and ninety degrees turns. The proposed approach results in a fast convergence algorithm able to noticeably reduce the number of objective function evaluations needed to solve this problem.