Water, as everybody knows, is a limited and fragile resource. In fact, water faces several challenges in the future due to the increase of population, pollution, climate change, etc. To overcome these problems, some concepts linked to sustainable development such as “Industrial Ecology” have emerged. The concept of industrial ecology is defined as a group of companies, close in proximity to each other, who cooperate with the objective of sharing resources such as energy, water or services in order to act in a more efficient way and more respectfully to the environment. However, environmental impacts of such organizations are still difficult to identify and the impacts on the local environment could be very different from a larger surrounding ecosystem such as a watershed. The aim of this work consists in defining the optimal configuration of EIP in the watershed so that impacts of industrial activities are minimized considering a large-scale system. This study develops a multiobjective optimization based approach by combining two models. The first one consists of the optimisation of the water networks of clusters of companies to make them work as eco-industrial parks where a multi-objective optimisation approach is used. Secondly, the impact of this ecological strategy/policy is studied at a bigger scale: the Garonne watershed, in France modelled through a Material Flow Analysis (MFA). This model helps to track and measure the quality of the water, before and after the implementation of the eco industrial parks. In each scenario studied, there is a significant reduction of the pollutant concentration, proving the benefits of the implementation of an industrial ecology policy at a larger scale. Depending on which industrial area is transformed into an EIP, one can observe a diminution of the pollutant concentration up to 70% at the estuary.