Ionosonde observations have been made at Great Wall station (62.22°S; 58.97°W), King George Island, and at further south Vernadsky station (65.25°S; 64.27°W), Argentine Islands, for many years. For several days at the two locations the magnetic meridional component of the thermospheric neutral wind has also been derived using three different algorithms with ionosonde data input. At King Sejong station (62.22°S; 58.78°W), close to Great Wall, almost simultaneous thermospheric winds were measured with a Fabry–Perot Interferometer (FPI) during a few days in 1997. All days correspond to intervals of low solar and geomagnetic activity levels and for different seasons. Here, the geographic meridional FPI winds measured at the geographic south pointing location are compared with the magnetic meridional component of the wind derived from ionosonde observations at Vernadsky. Also, the magnetic meridian FPI winds measured using all four cardinal pointing locations are compared with the magnetic meridional component of the wind derived from ionosonde observations at Great Wall. The patterns of the diurnal variations of the magnetic meridional component of ionosonde derived winds using the three different techniques are similar in most cases. However, the amplitudes of these variations and some individual values can differ by more than 150m/s depending on season, particularly during daytime. Comparison of the autumn FPI with the ionosonde winds for Vernadsky and Great Wall shows that they coincide within observation uncertainties. Results for other seasons are not so good. Some of the discrepancies are discussed in relation to the hour-to-hour variability of ionosonde based winds and the latitudinal gradients of ionospheric characteristics. Other discrepancies need to be further explained. Recently reported FPI mean winds for tens of days in different seasons for Palmer (64.77°S; 64.05°W), Anvers Island, are found to be particularly close to ionosonde derived mean winds for Argentine Islands station (former Vernadsky), albeit observations are for different time intervals. Unless comprehensive FPI winds for the Antarctic Peninsula longitude sector become available, ionosonde based winds seem to be reliable enough for several purposes.