We have modified an earlier model of Io's frozen surface SO2 and its influence on Io's atmosphere <Fanale et al., 1982>. In our new models, we consider four additional factors: (1) the latent heat of SO2 frost, (2) the rotation rate of Io, (3) thermal conduction and Io's internal heat flow, and (4) the deposition of some solar energy below the surface. The net result is a significant lowering of estimated surface temperatures and hence equilibrium SO2 pressures over the frost deposits. The atmospheric pressures are predicted to be much lower than those in the 1982 model of Fanale et al. and, in some cases, are low enough to be below the current upper limits set by recent Hubble Space Telescope observations. In the absence of local volcanic sources, even if Io were completely covered with frozen SO2, the atmosphere would be everywhere ballistic except very close to the subsolar point, where it would be marginally kinetic. Local kinetic atmospheres result only from SO2-rich plumes. The results of these models suggest that the sometimes reported posteclipse brightening of Io could be a real and temporally variable phenomenon and not an observational aberration. ¿ American Geophysical Union 1996