This paper describes a system dynamics model developed to facilitate design of nutrient reduction trading (NRT) programs. NRT is a form of cap-and-trade policy that has been strongly promoted in recent years to address diffuse source nutrient pollution. Despite its wide appeal and the enthusiasm of its proponents, very few trades have occurred to date. We propose that impediments to learning encountered in real-world pilot studies have contributed to lack of consensus in design and implementation of NRT. The model we offer is intended as a demonstration of how the design process in this instance can benefit from system dynamics modeling.