Total ozone relationships with selected upper troposphere/lower stratosphere variables (400‐ and 70‐hPa temperatures, tropopause height and temperature, 70‐hPa geopotential height, and 340‐K potential vorticity), as well as between the variables, are analyzed on decadal scales over Southern Hemisphere midlatitudes for the period 1980–2000. Total Ozone Mapping Spectrometer version 8 total ozone and European Centre for Medium Range Weather Forecast ERA‐40 data products for June and October (early winter and spring) are used. Multiple spatial correlation techniques and shared variance estimates are applied to infer relationships between mean fields as well as among decadal difference fields. Wave activity Z and local Eliassen‐Palm fluxes were calculated to further analyze the dynamics of the samples and their variability. The statistical studies show that observed total ozone latitudinal and longitudinal decadal variations can be driven by upper tropospheric and stratospheric variability, depending on latitude and season. The sampled regions, divided into subtropical and subpolar, yield differentiated relationships. October ozone decadal variations during the 1980s, particularly at higher latitudes, are attributed primarily to chemical ozone depletion, while there appear to be links between tropospheric decadal change and some of the stratospheric variables and tropopause behavior. In the 1990s, tropospheric contributions decrease, and stratospheric quasi stationary wave 1 plays a major role. In June, tropospheric change/variability appears to be more important than stratospheric driving, which nevertheless also contributes to change. Ozone change in the 1990s responded more to stratospheric dynamic change at higher latitudes, but despite reduced contributions, the troposphere remains a driver of variation at the lower latitudes of the sample.