A continuous simulation approach is proposed for estimating water temperature exceedance probabilities using thermohydrodynamic modeling. The approach uses (1) a deterministic unsteady flow and heat transport model, (2) continuous hydrological and meteorological data for a long historical period, and (3) synthetic records of tributary water temperatures and other model inputs. Representative historical records of streamfiow, air temperatures, and other hydrometeorological variables are obtained from nearby gages. Stochastic modeling methods are used to construct synthetic records for other model inputs, including inflow watertemperatures. An application of this deterministic-stochastic approach is presented for a complex waterway in northeastern fllinois with heat discharges from several power plants and wastewater treatment plants. Statistical results from the continuous simulations are compared to results obtained from traditional event simulations. The application illustrates the information that engineers and biologists can obtain for (1) evaluating compliance with water temperature standards, and (2) assessing the effect of water temperatures on aquatic habitat.