Why are formal statistical methods for risk-based decision-making so seldom used in the practice of watershed management? I contend that complex formal methods, while internally consistent, are often inappropriate to real world decision-making. The primary purpose of risk analysis is to support risk management, and decision methods need to be effective not just in evaluating risk, but also in communicating risk among stakeholders and decision makers. Useful methods must be not only correct, but also readily communicable. Many formal risk-based decision methods have real obstacles to practical application in one of the following areas: (1) many important components of risk that matter to stakeholders are difficult to express in quantitative terms, and any method which turns “fuzzy” information and subjective opinion into hard numbers is prone to be regarded with suspicion; (2) methods which are not understandable and convincing to decision makers have little practical value; (3) a complex formal analysis will be seen as misguided or irrelevant if it does not represent the full spectrum of management goals. This paper compares the process of watershed management with the process of ecological risk assessment, highlighting similarities and key differences. A practical decision method which balances quantitative rigor with ability to communicate to and forge consensus among stakeholders is then outlined with reference to a successful case study.