We developed a decision-support system for investment projects in pollution abatement plants at state-owned copper smelters in Chile. We formulated a mixed-integer linear model to determine the optimal investment policy for this interrelated production-environmental problem. Because of the significant differences between the model's solutions and the initial cleanup plans the smelters submitted to the environmental authorities, the government has revised the smelters' clean-up plans, initially working to obtain more complete and reliable information. Furthermore, it has put some of the projects presented in the original clean-up plans on hold until it has a refined version of the model. The Chilean budget director has approved other clean-up plans because, as the director of policy planning at Cochilco said, they are consistent with the model's recommendations.