A mesoscale, real-time, four-dimensional data assimilation (RT-FDDA) and short-term forecasting system has been developed for the U.S. Army Test and Evaluation Command (ATEC) at Dugway Proving Ground (DPG) in Utah. The DPG RT-FDDA system, which has been operational since the summer of 2000, is designed around the Pennsylvania State University/National Center for Atmospheric Research (PSU/NCAR) MM5 model. It uses a cycling methodology which provides the end-user with a continuous series of updated 3-dimensional analyses and a new 12-hr forecast on 3 domains (30 km, 10km, and 3.3 km grid spacings) every 3 h in real-time. The multi-stage cycling methodology is designed to take best advantage of the differing data ingest lags encountered in a real-time system (valid time lags can range from only a few minutes to 2 h), and uses a model-restart capability to provide continuous, balanced, 3-dimensional analyses and forecasts at any time interval. The FDDA system is a continuous assimilation system (Newtonian relaxation), and uses many diverse data sets, with availability frequencies ranging from 10 minutes to 12 h: standard surface and upper air observations, special surface observations, the University of Utah MesoWest observations (extensive western U.S. mesonet), DPG local surface observations, DPG profilers and RASS instrumentation, and NESDIS GOES wind vectors. The RT-FDDA system will be described, from design to product availability to computational performance, and both qualitative and quantitative verification results will be presented.