I compare satellite-derived global relief data on 13 seamounts in the northeast Pacific Ocean to echosounder-derived multibeam values from the Gulf of Alaska Seamount Expeditions and the West Coast Seamounts and Ridges Multibeam Surveys for peak height, latitude, longitude, and midsection area. I find Smith and Sandwell's (1997) global sea floor topography relatively accurate for peak geoposition but generally deeper than multibeam by 192 m (¿132 m). Cell size alone can explain 50--90% of peak height variation, suggesting higher-resolution geodetic altimetry could improve the accuracy of these estimates. Nevertheless, midsection areas overlap by 80 ¿ 11%. The altimetric technique clearly resolves the presence and general form of numerous unnamed deep seamounts, though predicted seamount peak height estimates for small features may vary from their true depth by up to 50%. These findings support high-end estimates of global seamount abundance since small seamount features (~1200 m relief) revealed by satellite may, in actuality, be quite substantial features (~2500 m relief).