Detailed shortwave and longwave radiative transfer models are used to calculate the radiative forcing and temperature trends due to stratospheric ozone depletion. These were calculated using the fixed dynamical heating approximation to adjust the stratospheric temperatures. Recent estimates of stratospheric ozone loss between 1979 and 1991 (from solar backscattered ultraviolet (SBUV) and stratospheric aerosol and gas experiment (SAGE) instruments) and updated radiative transfer schemes are used to obtain improved estimates of the radiative forcings. An annually and globally averaged radiative forcing of -0.13¿0.02 W m-2 decade-1(-0.22¿0.03 W m-2 for the 1979--1996 period) was found from SBUV total column ozone trends, applying a constant percentage ozone depletion to a 7 km thick layer directly above the tropopause. SAGE ozone trends gave forcing estimates of -0.10¿0.02 W m-2 decade-1 (-0.17¿0.03 W m-2 for the 1979--1996 period), although assumptions needed to be made about the choice of the vertical profile of the ozone depletion below 17 km. Using Dobson instrument trends from 1964 to 1996, the total ozone forcing could be as negative as -0.26¿0.05 W m-2. The quoted error bars derive from uncertainties in the total ozone trends. Using these values, the stratospheric ozone change may have offset about 30% of the forcing due to increases in well-mixed greenhouse gases since 1979, and about 15% of the forcing since 1964, at least on a global and annual mean. The ozone forcings are shown to be nearly a linear function of the ozone amount depleted from the atmospheric column, provided the vertical profile of the depletion remains constant. As in previous studies, it was found that stratospheric adjustment altered the sign of the ozone forcing by strongly cooling the lower stratosphere. It is shown that (depending on the vertical structure of ozone depletion) coolings of up to 0.4 K decade-1 can be found at altitudes of 35 km; this provides a mechanism for cooling the stratosphere nearly as large as that from well-mixed greenhouse gas increases, at altitudes where ozone changes were previously thought not to strongly affect stratospheric temperature trends. Previous studies have examined the effect of ozone trends by looking at the response of the surface temperature to a fixed absolute ozone change at different heights in the atmosphere. We argue that using absolute ozone perturbations places an unrealistically large emphasis on ozone changes near the tropopause. As the vertical profile of ozone change is more often reported as a percentage change, we present the sensitivity of the surface temperature to constant percentage, rather than absolute, changes in ozone.¿ 1997 American Geophysical Union