Vehicle emissions contribute to various greenhouse gas. About more than 80% of urban air pollution results from the transportation sector, according to a report of United Nations Economic and Social Council (2009). The transportation sector is the largest consumer of petroleum products and the chief culprit for air pollution worldwide (Fan and Lei, 2016; Zhou et al., 2017). The growth of energy consumption in transportation is faster than any other sectors due to the strong transport demands arising from rapid economic development. Therefore, green logistics in urban areas has been an important issue. This paper focuses on the green pickup and delivery problem (GPDP), an extension of the green vehicle routing problem (GVRP), which provides logistics service between pickup and delivery points. Pickup and delivery problem (PDP) is a famous extension of the vehicle routing problem (VRP). Practical applications of PDP include the dial-a-ride (Cordeau, 2006), airline scheduling, bus routing, and share-ride-problem (Caulfield 2009). We consider cooperation as a potentially powerful strategy to reduce carbon emissions due to its effectiveness in improving the utilization of vehicles. In a cooperative GPDP, the objective is to minimize carbon emissions; while each partner company involved in the cooperation has its own objective that is to minimize its cost. The cost of a company with a cooperation strategy may be more than that without a cooperation strategy. As such, the cooperation targeting on the lowest carbon emissions may not be accepted by all partners. The contradiction between the environmental target and the economic benefit calls for a reasonable compensation mechanism to form a feasible cooperative GPDP. Unfortunately, such a compensation mechanism for the cooperative GPDP has not been examined in the current literature. Compensation means that the cooperation can ensure a non-negative profit. Further, a method that can fairly distribute the profit obtained by the cooperation and compensation needs to be developed.