The Weather Research and Forecasting (WRF) model is quickly becoming one of the most widely used mesoscale meteorological models within the U.S. and throughout the world. The Department of Atmospheric Sciences at the University of Arizona has been using this model to perform real time forecasts during the monsoon season for the past few years. This presentation will discuss some of the recently funded projects in our department that are using this model in emerging research applications for investigation of the North American Monsoon. For improvement of short-term weather forecasts, an adjoint version of WRF is being used to determine forecast sensitivity to initial conditions for cases of severe weather events. A demonstration of the use of the WRF adjoint for a 2005 severe thunderstorm test case shows high sensitivity to the specification of a low-level moisture surge from Gulf of California and upper-level winds over the Sierra Madre Occidental. WRF will also be used as a regional climate model (RCM). It will be evaluated whether current summer seasonal forecasts can be improved by dynamically downscaling twenty years of retrospective global climate model data from the National Center for Environmental Prediction. A similar procedure will be used to dynamically downscale select IPCC model scenarios to potentially improve climate change projections of the monsoon. These RCM simulations will employ a spectral nudging technique, only recently developed for WRF, to ensure retention of large-scale variability.