Preferential escape of H relative to D will cause an enrichment of the atmospheric abundance of the latter over geological time. Measurement of atmospheric HDO and H2O provides an estimate of this enrichment (Owen et al. 1988). By assuming an initial H/D ratio similar to the Earth's, Yung et al. constructed a photochemical model based on the current atmosphere to determine the relative escape rates; they conclude that regular exchange of atmospheric water with a nonatmospheric reservoir containing about 3 m of water was required and that all but 0.2 m had escaped to space. This reservoir size appear to be inconsistent with current theories for evolution of the polar deposits, which suggest that they contain the equivalent of tens of meters of water and that this water exchanges on time scales which are short compared to the age of the planet. Solar forcing throughout the obliquity cycling may cause the time-averaged atmospheric water content to be larger than the current value. Significantly increasing the escape rates and increasing the required size of the nonatmospheric reservoir to a value consistent with the content of the polar deposits. If an early, wet atmosphere is present, as discussed by Owen et al, the requirements for polar cap exchange are eased. It is premature to conclude that the escape theories are inconsistent with polar cap behavior; better observations and models may, in turn, constrain the content and history of the polar caps. The models do suggest, however, that at least 60 m of water has escaped from the surface and atmosphere to space over geologic time. ¿ American Geophysical Union 1990