We have looked for the expected anticorrelation between H2O and O3 in the lower mesosphere using data from HALOE and a simple one-dimensional photochemical model. We use the fractional difference between the model and the data (ΔO3/O3) as a measure of the goodness of the fit of the model to the data. The model is run for each of over 1000 days of HALOE measurements from 30¿ S--30¿ N and from 1992--1999. Despite the 10% changes in upper stratospheric H2O which were observed during this period, we detect no significant ozone change in response to this H2O variability. Indeed, in the 2.0--0.4 mb region, the model error is lower when the water vapor is arbitrarily assumed to be invariant. Further, when H2O variability is considered, (ΔO3/O3) anticorrelates with H2O but not the temperature and solar flux. From this we conclude that the temperature and solar flux dependence are fairly well modeled and that the model error is due to the representation of the ozone dependence upon H2O in the model. Our results suggest that ozone near the stratopause is surprisingly insensitive to variations in H2O. If true, this insensitivity may make easier the observation of ozone recovery in the upper stratosphere due to declining chlorine. This result should also be considered when assessing the response of upper stratospheric ozone to increases in stratospheric humidity.