The number of severe weather events recorded in the national severe weather data base (Storm Data) has increased substantially over the last 30 years. This increase has been influenced by a number of factors, including: implementation of the national warning verification program, enhanced development of storm spotter networks, deployment of the national NEXRAD radar network, population increase and resultant growth of urban structures into previously rural areas, and an overall increase in weather awareness by some segments of the population. However, these effects have not been uniform across the U.S. resulting in regional biases and inconsistencies in the data base. Weiss and Vescio (1998) documented aspects of the reporting trends from 1955-1996, noting that the primary increase in severe hail, wind, and tornado reports occurred at the low end of the report spectrum (especially marginally severe hail and F0 tornadoes). They also found that severe wind reports were most difficult to classify and evaluate, because most reports consisted of various subjective degrees of "wind damage", or were listed as "peak wind gusts" such that measured or estimated values which were indistinguishable. The number of severe wind reports have increased dramatically (over 400%) during the past 30 years. In addition, the percentage of wind reports that contain a "peak wind gust" have increased from around 28% of all wind reports to more than 50%. Although there has been an increase in surface mesonetworks over the last decade, it appears that most "peak wind gust" speeds are estimated values and not obtained from calibrated anemometers. An examination of "peak wind gust" values at 1 kt intervals between 50 and 100 kt shows pronounced regional biases and discontinuities along various geographic borders and between areas under different NWS warning and forecast responsibility. In addition, it becomes readily evident that specific wind speeds (such as 50 kt, 52, kt, 60 kt) are favored in particular geographic regions, further contributing to sharp spatial discontinuities in the wind report data base. We believe that it is important to raise the level of awareness about the strengths and limitations of the severe wind event data base, especially in lieu of dramatic changes that have occurred since the middle 1990s. We acknowledge that operational challenges exist which have impacted the compilation of the data base, but it is important to stress the need for reliable and accurate reports so as to serve the needs of many users, including those who strive to conduct meaningful scientific research. Recommendations for improving the quality of the data base are presented, including ways to distinguish between different levels of reliability contained therein.