The development of methods to forecast deep convection and severe local storms in the United States is described from the 1940s until the present. This description includes the early efforts of the U. S. Air Force and the National Weather Service (NWS) Regional Forecast Offices as well as the development of forecast methodology within the NWS's centralized National Severe Storms Forecast Center (NSSFC) and Storm Prediction Center (SPC). The incorporation of research results into forecasting methodology and the resulting evolution from empirical forecast tools to a more scientific foundation (e.g., ingredients based) are discussed. Next, the current state of severe weather forecasting is described and examples of the forecast process are shown, including the role played by numerical weather prediction models. Limitations in our understanding of scale interactions and mesoscale/stormscale processes, and the resultant challenges we face to improve forecasting of deep convection and severe local storms are identified. Finally, suggestions are made as to how some of these challenges might be met.