Water quality trading (WQT) is a market arrangement in which a point-source water polluter pays farmers to implement conservation practices and claims the resulting benefits as credits toward meeting a pollution permit. Success rates of WQT programs nationwide are highly variable. Most of the literature on WQT is from an economic perspective emphasizing property rights, transaction costs, and the role of the price signal, but this institutional approach cannot fully account for the differential success rates between and within WQT programs. In this paper I turn to the role of social embeddedness and trust in helping explain these variable outcomes. Using interview data from all WQT programs nationwide I demonstrate that differential success rates between, and particularly within programs is largely determined by the presence or absence of a trusted local intermediary with relationships to the farming community. Trust helps WQT programs succeed because it makes information more credible to farmers, makes outreach to farmers more efficient, and buffers the risk that farmers may perceive in participating in the trading scheme. (Contains 4 tables and 3 figures.)