An air parcel trajectory model with chemistry initialized using data obtained by the Halogen Occultation Experiment (HALOE) has been demonstrated to be an effective tool for studying chemical and transport processes in the lower stratosphere. One of the sources of the differences between the trajectory calculations and the in situ observations from instruments aboard the NASA ER-2 aircraft is the uncertainty in the initial HALOE data. We compared HALOE ozone (O3) with near-coincident O3 obtained by the NOAA dual-beam UV-absorption photometer during the 1994 Airborne Southern Hemisphere Ozone Experiment/Measurements for Assessing the Effects of Stratospheric Aircraft campaign in a previous study. We have extended this study to obtain a more reliable evaluation of the differences in the remote and in situ O3 data by increasing the statistical base and using a revised version of HALOE data. In the present study we have extended the comparisons of O3 data to encompass four separate ER-2 airborne campaigns that have occurred since HALOE became operational in 1991. By extending the previous study to encompass the time span from 1991 through 1996, a broader range of latitude in both hemispheres, and a variety of seasons, we believe the conclusions resulting from the comparisons to be more robust and reliable. Statistical analyses show that the best agreement between HALOE and ER-2 O3 occurs in the 50--70 mbar region, where the difference of the means and RMS differences between the data sets are less than 7% and 28%, respectively, and where the HALOE data have an estimated error greater than 18% but less than 30%. Differences are larger in the 70--100 mbar region, where the maximum mean and RMS differences are 20% and 50%, respectively. These differences are somewhat less than those determined in the previous study. ¿ 1999 American Geophysical Union