The reverse logistics network (RLN) design for sustainable supply chain management is a strategic decision in network configuration, and is higher influenced by uncertainty. This paper applies a bi-level stochastic multi-objective model to design an RLN for a disposable product recycling management system. The goal is to balance the overall network cost against the associated environmental risks. An LP-metric based sample average approximation is formulated to solve the optimization problem. The model is validated numerically through a disposable product firm.