In this study, a linked simulation–optimization approach is proposed for optimal dewatering system design for an excavation site. In the simulation part of the proposed approach, an analytical groundwater flow model is used to simulate the flow process in the subsurface. This model is then integrated with two different optimization models where differential evolution (DE) and harmony search (HS) optimization algorithms are used. The objectives of both models are to determine the best dewatering system configuration by considering locations and pumping rates as decision variables. During the search process, the constraints related to drawdown and well locations are included in the optimization by using the penalty function approach. The performance of DE- and HS-based solution approaches is evaluated on a hypothetical excavation site for different well numbers. Identified results indicated that the DE-based solution approach provides slightly better results than those obtained by using the HS-based solution approach.