Ground‐based microwave techniques have supplied the only long‐term measurements of water vapor in the mesosphere. We review the entire current data base, which consists of measurements obtained in three separate experiments over an 8‐year period. The data from all three experiments indicate that the water vapor seasonal variation at mid‐latitudes in the upper mesosphere is dominated by an annual component with low mixing ratios in winter and high mixing ratios in summer. This suggests that the vertical distribution of water vapor in the upper mesosphere (up to 80 km) is controlled by advective rather than diffusive processes. This is consistent with the low mesosphericKzzvalues (≈ 105cm2s−1) deduced from the vertical gradient of the microwave water vapor measurements by Strobel et al. (1987). However, it is difficult to reconcile the predominantly annual water vapor variation with the semiannual variation in ozone at 78 km observed by the Solar Mesosphere Explorer. We perform a series of one‐dimensional photochemical/vertical transport model calculations which verify that (within the context of the hydrogen/oxygen chemistry considered in the model), the seasonal variation of water vapor cannot be the mechanism for the semiannual ozone variation. This variation is either a manifestation of some heretofore unknown ozone photochemical mechanism, or it could be driven by a seasonal variation in the vertical transport of atomic oxygen from the thermosphere. One possible vertical transport scenario for producing the semiannual ozone variation (while retaining the annual water vapor variation)