Availability of continuous long-term measured data for model calibration and validation is limited due to time and resources constraints. As a result, hydrologic and water quality models are calibrated and, if possible, validated when measured data is available. Past work reported on the impact of the length of the calibration period on the values of sensitive parameter values and performance of the Agricultural Policy/Environmental eXtender (APEX) model to simulate streamflow, but not water quality. This study used a long-term, continuously measured data from the Rock Creek watershed, located in northern Ohio, to quantify the impact of length of calibration period on sensitive model parameters and performance of APEX to simulate total nitrogen (TN) and total phosphorus (TP). Five short term (5-year), two mid-term (15-year), and one long term (28-year) calibration period scenarios were implemented. The calibrated parameters for each of the scenarios were applied to the validation period (last 3 years). The most sensitive streamflow, TN, and TP parameters and their values will be presented and discussed in detail, as well as model performance results during the calibration and validation periods. Finally, computed uncertainty ranges of simulated streamflow, TN, and TP outputs will be also be presented. These results will be used by the USDA to develop an APEX model parameter database to support nation-wide deployment of the nutrient tracking tool (NTT), a user-friendly web-based computer program used to estimate reductions in nutrient losses associated with alternative practices.