Market-based approaches to address water quality problems have resulted in only limited success, especially in trading programs involving both point and nonpoint sources. We analyze one of the largest point-nonpoint trading programs - the Great Miami Trading Program in Ohio, administered by the Miami Conservancy District (MCD). Our evaluation focuses on the economic and institutional aspects of the program, including cost effectiveness, efficiency of bidding, transaction costs, trading ratios, and innovation. We use a unique dataset consisting of all bids from agricultural nonpoint sources and interviews of soil and water conservation district (SWCD) agents in the watershed. We find that the MCD's reliance on county-level SWCD offices to recruit and advise farmers has been essential to achieve relatively high rates of farmer participation. Additionally, the MCD is able to partly ride on the administrative costs that SWCD offices receive to assist federal conservation programs, which is helpful to lower costs for a fledgling trading program. However, the involvement of SWCD offices reduced the potential cost savings from the reverse auction structure because some agents were able to learn about the threshold price over the six rounds of bidding and help farmers bid strategically. Overall, the program structure serves as an effective model for future trading programs in other regions that seek to involve agricultural nonpoint sources.