The use of watershed models has become widespread, and the complexity and specificity of the model applications are ever increasing.  However, the basic need for modeling was to simulate hydrology and constituent transport processes at the field to watershed level – allowing for the model user to then evaluate the effects of various exogenous factors on hydrology and water quality.  Watershed models are cool and sexy, and this aspect has those not informed about model uncertainty trusting the output to make management and maybe even regulatory decisions within our TMDL framework.  The environmental important Illinois River Watershed in Arkansas provides the case study for this presentation, where two different players were prioritizing subwatersheds (HUC 12s) for a non-governmental organization’s watershed based management plan and a state agency’s target for BMP implementation funds.  This controversy lead to two state agencies requiring that water-quality monitoring and modeling be conducted jointly in the other HUC 8 watersheds requiring prioritization.  A couple techniques emerged to evaluate model output in the absence of sufficient calibration data using monitoring results from outside the modeling period.  This poses the question – if you need the monitoring data to validate the model output, then why not just use the monitoring data to guide management decisions?