Multiple regression methods are applied to estimate the solar cycle variation of (1) zonal mean ozone as a function of altitude and latitude using a combination of Nimbus 7 solar backscattered ultraviolet (SBUV) and National Oceanic and Atmospheric Administration (NOAA) 11 SBUV/2 ozone profile data for a 15-year period; (2) total ozone as a function of latitude, longitude, and season using Nimbus 7 total ozone mapping spectrometer (TOMS) data for a 13.3-year period; (3) lower stratospheric temperature as a function of latitude, longitude, and season using microwave sounding unit (MSU) Channel 4 data for a 16-year period; and (4) lower stratospheric geopotential height as a function of latitude, longitude, and season in the northern hemisphere using Berlin height data for a 30-year period. According to the SBUV-SBUV/2 data, most (about 85%) of the 1.5--2% solar cycle variation of global mean total column ozone occurs in the lower stratosphere (altitudes <28 km). Evidence is obtained for a related solar cycle variation of lower stratospheric temperature (50--150 mbar) and geopotential height (30, 50, and 100 mbar) with geographic dependences similar to that of the solar cycle variation of total ozone. Specifically, total ozone, lower stratospheric temperature, and lower stratospheric geopotential height have annual mean solar regression coefficients in the northern hemisphere that reach a maximum near 30 ¿N latitude within a longitude sector extending from approximately 160 ¿E to 250 ¿E. Maximum variations from solar minimum to maximum in this sector are approximately 11 Dobson units, 0.8 K near 100 mbar, and 60 m at 50 mbar, respectively. Seasonal solar regression coefficients tend to be statistically significant over larger areas in summer but have larger amplitudes within limited regions in winter. These geographic similarities between total ozone, lower stratospheric temperature, and geopotential height solar coefficients suggest that changes in lower stratospheric dynamics between solar minimum and maximum may play an important role in driving the observed total ozone solar cycle variation. To test this hypothesis, a simplified perturbation ozone transport model is applied to calculate the expected total ozone variation owing to dynamical forcing for the calculated geopotential height solar coefficients, climatological ozone mixing ratios, and zonal winds. For the summer season during which the solar regression coefficients are significant over the largest area, both the amplitude and latitude dependence of the observed solar cycle ozone variation are approximately consistent with the model estimates.¿ 1997 American Geophysical Union