A simple statistic is derived for quantifying the potential for the aliasing of tidal errors in a given linear estimate of sea surface height constructed from altimeter data. The existence of M2 tidal constituent errors in Geosat data processed in the traditional way (i.e., with orbit errors removed using least squares fits to 1 cycle per revolution sinusoids) which are of sufficient magnitude to alias into apparently westward propagating ocean features is demonstrated by artificially inducing aliasing. The aliasing statistic presented here responds clearly to the induced aliasing and to actual aliasing caused by real data dropouts in the Geosat data. The potential for aliasing M2 tidal errors is shown to vary with latitude depending on the time interval between ascending and descending ground tracks near the location of interest. The methods developed here are applied to Geosat data from the northeast Atlantic to demonstrate the presence of M2 tidal error aliasing in those data. ¿ American Geophysical Union 1994