A model has been developed that takes weekly stream temperature and weather data and determines the potential effects of stream channel shading, wind sheltering, and secondary (hydrologic) sources of heat on the observed stream temperatures. The partitioning (parameterization) of these effects is done via an empirical optimization procedure known as the shuffled complex evolution global optimization method. On the basis of data from 596 stream gauging stations in the eastern and central United States, the results of this analysis indicate that, on average, the effect of shading is stronger than the effect of sheltering and that the inclusion of a parameter for secondary heat input improves the simulation of the stream temperatures. The results increase our understanding of annual and seasonal controls on stream temperature at a regional spatial scale. By providing the above information for a large number of streams in the eastern and central United States, we have learned about typical values and ranges of stream shading and sheltering and have learned that hydrologic heat inputs cannot be ignored for many stream reaches. We have developed more realistic expectations of secondary heat inputs to streams. With this information, we will be better prepared to estimate and appreciate the effects of land use changes or climate changes on stream temperatures and hence on stream habitats.