The National Weather Service (NWS) issues many thousands of warnings each year, and these warnings vary spatially and temporally. This study served to quantify severe thunderstorm and tornado warning differences from 2005-2012 for the 116 NWS offices in the conterminous U.S. This was investigated by looking at regional differences, as well as changes occurring with the switch from county-based warnings to storm-based warnings on October 1, 2007. It was found that length and lead time of warnings were variable per region, and the change in lead time from county-based to storm-based warnings was negligible for most Weather Forecast Offices. The switch from county-based to storm-based warnings impacted the size of warnings most particularly in the Western Region of the NWS. Measures of forecast quality such as Probability of Detection and Success Ratio improved for most offices with the transition from county-based to storm-based warnings. The scope of this work did not include any societal aspects of warnings, such as whether the warning is received by the emergency managers and/or the public, but this is an important factor in warning success that is not considered here.