The 10-day World Day campaign during January 20--30, 1993, provided an opportunity to test the current capability of the National Center for Atmospheric Research general circulation models and to conduct simple numerical experiments to investigate possible causes of day-to-day variability. Detailed data sets from the Arecibo and Millstone Hill incoherent scatter radars provided information on the middle- and low-latitude thermosphere and ionosphere during low solar activity which can be compared with model predictions. The theoretical model was used to examine the impact of varying two of the model inputs: the high-latitude energy and momentum sources and the semidiurnal tidal waves from the lower atmosphere. These exercises indicated that varying the high latitude inputs affect the simulations even to relatively low latitudes. The neutral winds in the models were responsive to the level of auroral activity and also to the magnitude of the waves from the lower atmosphere, particularly the neutral zonal winds. The simulated hmax were only affected at night by varying the model inputs. Use of the assimilative mapping of ionospheric electrodynamics (AMIE) technique is necessary to produce realistic quiet-time zonal ion drifts at low latitudes following local sunset. The ion and neutral temperatures proved nearly insensitive to the specifications of the auroral or the tidal inputs, particularly the temperatures at Arecibo. This is in contrast to the observations in which temperatures may vary by up to 100 K from day to day with more pronounced variability at night. In the models, only a large geomagnetic disturbance produced a perturbation in the temperatures but with magnitudes significantly smaller than those observed. The discrepancies may indicate an underestimate of the high-latitude Joule heating due to small-scale variability in magnetospheric electric fields, which would affect the neutral circulation and composition, and inadequate representation of the F region dynamo and conjugate effects in the models.¿ 1997 American Geophysical Union