Radar observations in a tornado that occured during the 3 May 1999 outbreak in Oklahoma reveal Doppler velocities of approximately 135 ms-1. While these have been widely, and loosely reported as ?record high winds?, the proper interpretation of these and other radar measurements of tornadic winds is difficult.

Observations from 3 May 1999 will be presented in comparison with other high wind measurements made by the DOWs in both 1998 and 1999. Interpretation of radar wind observations will be discussed. The difficulties of comparisons with standard surface wind measurements, both temporally and spatially, comparisons with F-scale ratings, the complications caused by air vs scatterer measurements, and the differences between mean Doppler measurements and spectral measurements will be addressed.