Spill events are highly probabilistic occurrences. Major spills that can have very significant impacts on raw water sources of drinking water are relatively rare events. A systematic method for designing and operating early warning systems that considers the highly variable, probabilistic nature of many aspects of the system is needed. These include probability of spills, behavior of monitoring equipment, variable hydrology, and the probability of obtaining information about spills independent of a monitoring system. This paper describes the characteristics and application of Spill Risk, a risk-based model using Monte Carlo simulation techniques, that has been developed as part of an AWWA Research Foundation project. Includes 2 references, tables.