In this study, we compared several instruments designed to mitigate nonpoint source pollution on one given watershed. This comparison has been undertaken on three axis, the potential effectiveness of each instrument, the associated costs expressed as the variation of the farmers and the tax-payers' surplus, and the acceptability of the regulation assessed as the proportion of the farmers who benefit from this regulation, or who are ready to voluntary adopt it. The instruments that have been compared include technical modifications of the current farming practices, taxes or quotas on the amount of fertilisers used by the farmers, a linear extensification of the production, and non linear instruments. Non linear instruments are of special interest when the practices are not easily observed: in this case, the regulation of pollution can be designed on the basis of a small set of variables that are common knowledge (in our case, the level of production is easily observed). The subsidy scheme is based on a contract menu: each farmer chooses in this menu a contract that he believes is the best for him. This contract menu is designed such as optimising the social surplus (farmers' plus tax-payers' surplus) less the damage. The first result underlined by this study is that there are many ways to mitigate non point source pollution from farming activities on a watershed. Most of them have been simulated as effective enough to reach the EU threshold of 25 mg NO3-/l. The second result puts forward the difference between the feasibility of technical BMPs implementation and their acceptability by the farmers. Obviously there is a large place here to improve the environmental advice strategies. Third, the cost investigation suggests that optimally differentiated regulations are the best way to conciliate effectiveness, implementation costs and acceptability of mitigating instruments. In our application case, the loss of welfare related to private information is lower than the cost of information required to implement the first best production levels and obviously such a program with adjusted subsidies is cost-effective to implement the Water Framework Directive.