Recent studies of the solar cycle variation of ozone have shown that the response of ozone in the upper stratosphere to solar UV variation, as inferred from the SBUV (Solar Backscatter Ultraviolet) type measurements, is about a factor of two greater than estimated from 2-D photochemical models. Because of potential errors in accounting for the long term instrument drift in the SBUV type of measurements, the significance of this discrepancy is difficult to quantify. In this paper, ozone measurements from the Microwave Limb Sounder (MLS) and the solar irradiance measurements from the Solar Stellar Irradiance Comparison Experiment (SOLSTICE) and the Solar Ultraviolet Irradiance Monitor (SUSIM) onboard the Upper Atmosphere Research Satellite (UARS) are analyzed to estimate the upper stratosphere ozone response to changes in the solar UV irradiance. During the three year period of UARS measurements, analyzed here for the declining phase of the solar cycle 22, the solar irradiance in the 200--205 nm range decreased by about 5% from a near solar maximum to a near solar minimum level. During the same period, ozone mixing ratio measured from the MLS instrument decreased by about 2--4% in the 0.7--3 hPa region. In the upper stratosphere, the general characteristics of the MLS time, series are similar to those inferred from the NOAA-11 SBUV/2 measurements. The SBUV/2 trends above 1.5, hPa, however, are significantly greater than those derived from the MLS data. The UARS data suggest that the long term solar UV response of ozone in the upper stratosphere is underestimated by 2-D photochemical models as in previous studies based on the SBUV type measurements. ¿ American Geophysical Union 1996