Model performance assessment, also known as skill assessment, is a necessary and critical element in the application of watershed models for water resource and water quality management. A comprehensive assessment of model performance includes:1. Consideration of how well a model is able to simulate observed data that describe the watershed's hydrologic and water quality response to its forcing functions (e.g., meteorology, land disturbance activities, point and nonpoint source loadings), 2. Measuring the relative sensitivity of model output to various model parameters in the specific setting in which the model is being applied, and 3. Assessing the potential uncertainty that is introduced into model output as a result of naturally occurring variability in the actual values of model parameters.By means of model calibration and validation the modeler gains understanding of the response behavior of the watershed that is being modeled and the ability of the parameterized model to mimic that response. Sensitivity analysis and uncertainty analysis provide both the modeler and those who use the model results with a means of understanding a model's inherent strengths (or limitations) in accuracy. The combination of these individual assessments determines the level of confidence that model users should invest in a model's ability to represent current and alternative watershed responses.