ABSTRACT The fleet sizing problem is a critical challenge in closed-loop supply chains because of the additional requirements of managing product returns. To tackle this challenge, this study addresses the integration of fleet sizing with vehicle routing problem in closed-loop supply chain management, considering simultaneous delivery and pickup under demand uncertainty. A mixed-integer linear programming model is proposed to minimize vehicle acquisition and shipment costs, and a multi-stage adjustable robust optimization formulation is employed to handle uncertainty. Prior studies have neglected to address this important problem due to a lack of efficient solving methods. Hence, three exact, heuristic, and metaheuristic algorithms are developed to solve the problem. The numerical experiments demonstrate that the proposed methods outperform the existing Hyper-box based Branch-and-Partition method, improving computational efficiency and solution quality. This study also helps practitioners make better fleet sizing decisions under uncertainty.