Model calculation of ozone trends in the 1982--1986 time period show that interannual stratospheric temperature variations over that period dominate over changes in minor constituents, including CLx, in producing trends in ozone. Estimates for upper limits to changes in solar flux over the declining part of the solar cycle and estimates for maximum increases in Clx lead to similar model decreases in upper stratospheric ozone, although the calculated trends from each are less than 1%/yr. Model calculations suggest that long-term increases in Clx may produce decreases in the relative amplitude of the annual variation of ozone near 1.0 mbar than are nearly twice the secular decreases in ozone. However, trends in temperature over the period 1982--1986 lead to predicted decreases in the seasonal amplitude of ozone of up to 10%/yr, based upon National Meteorological Center (NMC) observations of upper stratospheric temperatures. The secular and seasonal ozone trends derived from SBUV data for the 1982--1986 time period exhibit poor agreement with model predictions. SBUV data indicate secular trends in ozone chemistry. SBUV data also exhibit significantly smaller changes in the amplitude of the annual variation of ozone than the models which include NMC temperatures. At least part of this disagreement may be related to temperature-correlated variations in photochemistry which are not included in the model, such as changes in water vapor or other minor constituent densities. Future efforts to determine ozone trends from satellite measurement should include monitoring of ozone, temperature, and possibly water vapor in the atmospheric pressure region from 10 to 0.1 mbar. ¿ American Geophysical Union 1988