Spaceborne radar altimeters make high‐precision measurements of the range between the satellite and the ocean surface. Atmospheric water vapor will lengthen the range as measured by an altimeter. It is theoretically possible for horizontal changes in atmospheric water vapor to be misinterpreted as changes in sea surface height (SSH) or to mask SSH changes that are present. Scanning multichannel microwave radiometer (SMMR) data from the entire 3‐month lifetime of the Seasat satellite are used to estimate water vapor path length corrections and to assess the probability that water vapor will significantly degrade the interpretation of Geosat altimeter data. It is shown that if one considers Geosat‐observed mesoscale circulation features of 10 cm or greater, particularly in the context of a regional circulation model, water vapor is not generally a problem. For mesoscale circulation features 5 cm or smaller, potential contamination by water vapor gradients becomes significant. Cumulative distributions describing water vapor statistics for a variety of spatial scales and in specific geographical regions, as well as the entire globe, are presented. These distributions are compared with ones derived using Fleet Numerical Oceanography Center (FNOC) predictions. It is demonstrated that FNOC‐derived water vapor path lengths, although providing a good mean correction, will not serve to compensate for small horizontal scale changes in water vapor.