During the first 6 days of March 1995, measurements of the ionospheric electron density were made with a digisonde, and thermospheric winds were measured with a Fabry-Perot interferometer. This was a period of low solar activity and moderate to high magnetic activity. The ionograms have been scaled and the traces inverted to obtain the electron density profile and the peak height of the F2 layer (). Modeling has been employed to derive equivalent thermospheric neutral winds at hmF2. The derived neutral winds are in very good agreement with the measured optical winds most of the time. The winds follow a strong diurnal pattern with poleward winds during the day, weak winds near dawn and dusk, and strong equatorward winds peaking near local midnight. On most nights the peak equatorward wind speed was around 200 m s-1, but on March 1 it did not exceed 110 m s-1. For these magnetic and solar activity conditions the wind at the F2 peak altitude (~350 km) from the HWM93 empirical wind model [Hedin et al., 1996> did not exceed 90 m s-1 at any time but was in generally good agreement with the hmF2 wind during the day and with both measured winds on the nights of March 1 and 2. The good agreement between the optical and hmF2 winds was obtained by using the recommended Burnside factor of 1.7 to multiply the O+-O collision frequency, but better agreement was obtained either by using a Burnside factor of 2.0 or by increasing the atomic oxygen density by 20%. Recent suggestions of much lower Burnside factors could be tolerated only if there were large systematic errors in the measurements or large electric fields. ¿ 1997 American Geophysical Union