This paper will examine impacts of accurate ozone data in the lower stratosphere on the meteorological analyses produced using a state-of-the-art data assimilation system. The Goddard Earth Observing System, Version 5 (GEOS-5) data assimilation system is used to produce analyses and forecasts of the atmospheric state, including the ozone distribution. The standard version of GEOS-5 uses a suite of observations, including microwave and infrared radiances from operational (e.g., AMSU) and research (e.g., AIRS) sensors, as well as ozone retrieved from Solar Backscattered Ultraviolet (SBUV) measurements. Although constraints on partial columns are very good, the SBUV data have little information content about the ozone profile between the surface and about 64hPa. Ozone retrieved from the Microwave Limb Sounder (MLS) instrument on NASA's EOS-Aura platform has information about the ozone profile from pressures as low as 200hPa, with a good signal-to-noise ratio in the lower stratosphere, meaning that the ozone profile can be well constrained in this region. The Ozone Monitoring Instrument (OMI) on EOS-Aura provides information on total ozone with a fine spatial footprint. This work will examine the impacts of using EOS-Aura ozone data on GEOS-5 analyses: specifically, MLS retrieved profiles and cloud-free OMI retrievals are used. The focusing of the presentation is on the causes of improvements in the upper troposphere and lower stratosphere that is indicated by reduced observation-minus-forecast residuals in the AIRS ozone channels and by changes in the circulation.