A time-dependent calculation has been performed to study the dynamical and compositional response of the thermosphere to auroral disturbances. The heat and momentum inputs associated with auroral activity were estimated by using an empirical model constructed from two data bases: the precipitating particles monitored by the TIROS/NOAA satellites and the ionospheric convection electric fields measured by the Millstone Hill incoherent scatter radar. The level of auroral activity was given in terms of hemispherical power input due to particle precipitation. The model of high-latitude energy input was incorporated into a zonally averaged thermospheric model. This paper presents the results of a simulation done for the period September 18--24, 1984. Time histories of the composition are given as well as those of dynamical parameters such as temperature and winds. The spatial structures in these parameters are also discussed. The main new result of the simulation is that during auroral disturbances the radiative cooling by nitric oxide (NO) is greatly enhanced by the increase in NO number density as well as by the temperature increase. As a result the temperature increase in response to auroral disturbances the radiative cooling by nitric oxide (NO) is greatly enhanced by the increase in NO number density as well as by the temperature increase. As a result the temperature increase in response to auroral activity is damped, and the temperature relaxation time is shortened following a quieting of activity. A comparison between the simulated temperatures and the temperatures observed by the Millstone Hill incoherent scatter radar during the 7-day interval shows reasonable agreement in terms of the absolute magnitude of the variations as well as the time constants of those variations. ¿ American Geophysical Union 1989