The transportation problem is commonly applicable in the field of operations research. It is a subclass of linear programming problem. The main objective in transportation problem is to minimize the cost of shipping homogeneous commodity from various origins to various destinations with respective rim requirements. The authors propose an extension to the classical transportation problem by considering multiple incommensurate inputs and outputs for each shipment link. The relative efficiency concept is defined for each shipment link. Two linear programming models are used to determine the optimal transportation plan with maximum efficiency. The applicability of the approach is discussed with a numerical example.