In the realm of e-commerce, various return freight modes are available, such as platform-bearing return freight (PP strategy), retailer-bearing return freight (RP strategy), and consumer-bearing return freight (CP strategy). The return freight strategy adopted by platforms plays a crucial role in responding to consumer behaviors. To analyze how platforms determine the optimal return freight strategy, we examine a supply chain comprising a platform, a retailer, and heterogeneous consumers, employing a Stackelberg game model to capture their decision dynamics. Our findings suggest that the platform opts for the PP strategy only when the product profit margin is relatively low. If the platform refuses to adopt the PP strategy, then the RP strategy is preferable when consumers face extremely high consumer return freight costs; otherwise, the CP strategy is the equilibrium result. We also prove that the selection of the return freight strategies leads to changes in the property of the equilibrium. Under the CP strategy, how the product fit probability affects the optimal price depends on the procurement cost; while under the RP and PP strategies, the optimal price decreases with the fit probability. Moreover, under the PP strategy, an increase in the product fit probability does not always benefit the platform. Further, our analysis reveals that, while the PP strategy can reduce the retailer’s risk, it might negatively impact supply chain performance, which implies that cost sharing in the supply chain does not necessarily lead to a higher channel profit. We also investigate how the platform’s return freight strategies affect consumer surplus and identify regions for Pareto improvement where all parties benefit from the PP strategy. Finally, by exploring extensions like positive salvage values, partial refunds, platform competition, and differentiated return freight costs, we confirm the robustness of our findings and unveil new insights.