Numerous aircraft, ship, buoy and land nation data were composited with respect to the center of Hurricane Frederic for two time periods: a 24 h period corresponding to the storm’s position in the open Gulf of Mexico on 12 September 1979, and an 8 h period corresponding to the landfall of Frederic near 0400 GMT on 13 September. Comparison of wind analyses for the two periods indicated a rotation of maximum inflow angles from the southeast to northeast quadrants and a strong frictional decrease of wind speed over land. Them and other features of the landfall analysis were compared with a model landfall study by Moss and Jones (1978). The landfall composite wind field was compared with the Fujita damage vector analysis to determine the damage time interval and mean wind speed range. Damage vector directions were found to be well correlated with the surface streamlines, with the most severe damage being associated with Frederic’s northern eyewall. Ten-meter-level wind speed data over water (VO) and at coasts stations (VL) were used to formulate approximate relationships of the low-level (500–1500 m) aircraft wind (Va) to the mean coastal wind and peak gust (VLG) in the same position relative to the storm center. It was found that VO = 0.7Va, VL = 0.8V0, VLG = 0.8Va and VL = 0.56Va. These relationships should aid from in their 2 assessments of low-level aircraft reconnaissance wind data for use in issuing warnings. The vertical shear of the horizontal wind determined from from radiosonde data for two inland stations was compared with shear determined from surface and data data over water. The overland shear was greater than the overwater shear, by a factor of 2, in the same relative part of the storm. The “thermal wind” shear computed in the vicinity of the center was negligible, although the 10 m level air temperature analysis over land indicated a cold core that was probably caused by adiabatic cooling.