The depletion of dissolved oxygen in coastal and estuarine waters has been increasing in frequency and intensity in recent decades. Hypoxia results from complex interactions between physical and biogeochemical processes and causes stresses in the marine environment. The cause, remediation and hopefully reverse of this widespread phenomenon has drawn many research interests. Its coincidence with eutrophication has also spawned management debates and has profound influence on management decisions. Well-tested numerical models are invaluable tools for studying, predicting and informing site-specific management decisions. As part of the U.S. IOOS testbed project managed by SURA, this effort focuses on applications and inter-comparisons of multiple models to address the seasonal hypoxic conditions in two large coastal ecosystems in the United States: the Chesapeake Bay and the northern Gulf of Mexico. In both systems, we implement Fennel's NPZD-type biochemical model in NOAA's Operational Forecast Systems (ROMS-based CBOFS for Chesapeake Bay and FVCOM-based NGOFS for northern Gulf of Mexico) and compare results with various other efforts in each region. Configuration differences in the two systems give insights into processes contributing to the seasonal oxygen depletion. The strengths and issues of different hydrodynamic models and different configurations of oxygen- regulating processes will be identified and discussed. The results will provide guidance in building up the ecological forecasting capabilities at NOAA.