The presence of end-of-life vehicles (ELV) in the cities creates irreparable damage in environmental and economic terms. Thus, governments have been searching for ways to collect ELVs. An effective way to address ELVs is subsidies' policy. In this study, a supply chain including an ELV take-back center, an inspection center, and a repair center are considered. The decision variables are the purchase price of the ELVs, the sale price of the repaired vehicle, and the level of vehicle repair. In this supply chain, the government pays a subsidy to take-back centers that deliver their used vehicles. A Stackelberg game structure is considered with the government as the leader, the inspection center as the primary follower, and the collection and repair centers as the second followers. Using the game theory approach, the effects of government subsidies on equilibrium values of the decision-making variables of the centers in the ELV supply chain have been investigated in three scenarios. The scenarios comprise profit sharing, revenue sharing, and a centralized (cooperative) scenario. The results indicate that the centralized scenario has the most significant advantage compared to the two contracts of profit sharing and cost sharing. For example, the repair level of the ELVs, the amount of repaired vehicles demand, consumer surplus, and the whole supply chain's profit are at their maximum. The most significant impact of subsidy payment is related to the take-back center's profit and the ELV supply. On the other hand, the subsidy does not affect the profit of the repair center.