We compare middle atmospheric water vapor measurements from the HALogen Occultation Experiment (HALOE), Water Vapor Mm-wave Spectrometer (WVMS), and Polar Ozone and Aerosol Measurements (POAM) instruments. Coincident WVMS and HALOE data are compared at the Network for the Detection of Stratospheric Change (NDSC) sites at Mauna Loa, Hawaii (19.5N, 204.4E), from 1996–2002 and at Lauder, New Zealand (45.0S, 169.7E), from 1994–2002. We find that, at most altitudes, there is no significant instrumental drift between the HALOE measurements and the measurements obtained from the two WVMS instruments over these time periods. We also find that at altitudes below $60 km, where the solar cycle effects are unimportant, there were no significant trends in either the HALOE or WVMS data sets over the periods of intercomparison. This is in marked contrast to the early 1990s where large trends (2%/year) were obtained from both HALOE and WVMS measurements. Similarly, we compare the HALOE data with the high-latitude 1998–2002 POAM data set at equivalent latitudes from 45N–55N. Again, the trends are small, and the instrumental trend differences are at most only slightly >1s. We also look at the HALOE water vapor and H2O + 2CH4 data globally and find that, after a large increase from 1991–1995, there is almost no trend below $60 km from 1996–2002. The water vapor trend over the entire HALOE measurement period (1991–2002) is <1%/year. The unusual stratospheric dynamics at the beginning of the 1990s caused large changes in upper stratospheric and lower mesospheric water vapor that were unrelated to the amount of water vapor entering the stratosphere, hence in this region the less dynamically sensitive quantity H2O + 2CH4 probably provides a better measure of water vapor entering the stratosphere during this period. The trend in H2O + 2CH4 over the 1991–2002 HALOE measurements period is $0.5%/year.