Targeting has become the buzz word in the national agri-environmental policy reform in Finland. It is generally accepted that more environmental benefits could be reaped by implementing environmental protection measures where they have the biggest positive impact. However, considering one of the main environmental problems in Finland resulting from agriculture, eutrophication, the identification of first-best policy or even the biggest contributors among the diffuse nutrient sources remains as a considerable challenge. The model developed in this study aims to demonstrate how the agricultural nutrient load potential can be calculated in a way which supports the identifying of cost-efficient abatement policies. We use metamodeling of dynamic nutrient load model (ICECREAM) to establish load parameters for non linear economic optimization to derive abatement cost functions for nutrient loads of 2 Finnish catchments. We calculate the difference in costs of the spatially optimal allocation of reduction measures and compare with the costs based on average non-targeted measures.