The thermospheric effects of geomagnetic history and the resulting ion-neutral interactions are determined through the analysis of a long-term database of high latitude neutral winds from Kiruna, Sweden, and simulations with the coupled thermosphere ionosphere plasmasphere model (CTIP). Three types of geomagnetic history are examined in detail with the data and the model: steady state conditions in which the Kp index for the current three hours is the same as the Kp index for the previous 3 hours; previously quiet conditions in which the Kp index for the current 3 hours is greater than the Kp index for the previous 3 hours; and previously active conditions in which the Kp index for the current 3 hours is less than the Kp index for the previous 3 hours. It is shown that during the hours of darkness at Kiruna, while the ionosphere responds immediately to changes in activity, the neutral gas can take between 3 and 6 hours to recover from the effects of any previous activity. Model simulations show that the rate of energy dissipation is also significantly dependent on geomagnetic history. For the previously active case the Joule heating and mechanical energy transfer rate are up to 4 times larger at certain latitudes as the steady state case. For the previously quiet case the heating rates are much smaller than the steady state case. There is a frequently made assumption that at high latitudes the mechanical energy transfer rate may be ignored as insignificant compared with the Joule heating rate. The results presented here show that this assumption is unreliable, particularly in the dusk sector and polar cap. ¿ 1999 American Geophysical Union