Prior to concerns of sustainability, almost all industrial engineering models tried to minimize cost or maximize profit. Sustainability awareness has recently forced the decision makers to also take into consideration such aspects as clean water use, or carbon dioxide emissions. In an effort to incorporate more aspects of sustainability in optimizing production efforts, we present a network model to handle trade-offs among dissimilar sustainability criteria. Since typically there are alternative choices for the various operations, the network allows parallel arcs between the same nodes. We also introduce the concept of generalized cost. Generalized cost is a vector quantity that includes not only a monetary measure, but also measures relevant to sustainability, such as carbon use or embodied energy. The approach leads to a multi-criteria decision making model, whose efficient frontier is obtained by the epsilon constraint method. Numerical work shows that the computational effort to obtain the efficient frontier is reasonable, allowing products of up to about a hundred activities to be solved with the current generation of personal computers.