Giant polygons on Mars with several kilometers in diameter were first observed in Mariner 9 and Viking Orbiter images, and their origin, formation, and evolution have remained enigmatic since that time. New data obtained by the Mars Orbiter Laser Altimeter (MOLA) and the Mars Orbiter Camera (MOC) on board the Mars Global Surveyor spacecraft now permit analysis of topography and morphology of the polygonal terrain in unprecedented detail. MOLA data show that (1) giant polygonal terrain in Utopia is located at the lower slopes of a basin structure, that is the Utopia basin; (2) the onset of polygonal terrain in Utopia is close to the same elevation in several profiles and lies at about the same elevation as a terrace which was interpreted as a shore-line of an ancient body of standing water within the Utopia basin proper; (3) polygonal terrain occurs over a wide range of elevations on regional slopes of ~0.1¿; (4) the depths of the troughs range from <5 m to 115 m, averaging about 30 m, and tend to be greater toward the center of the Utopia basin; (5) the mean width of polygonal troughs is of the order of 2 km, ranging from <0.5 to 7.5 km. Data from the Mars Orbiter Camera (MOC) indicate that (1) the troughs are generally broad graben-like features with varying morphologies (i.e., terraces and lobateness) and dimensions (widths and depths); (2) trough morphology has been modified by aeolian processes; (3) polygonal troughs are weakly interconnected at MOC resolutions and there are no smaller polygons observed within the giant polygons; (4) circular depressions which are probably related to subsurface collapse sometimes occur at the bottoms of polygonal troughs; (5) small-scale polygonal terrain on the ejecta blankets of young craters with fluidized ejecta in the investigated area exhibits different characteristics (i.e., smaller size, presence of rims, greater degree of interconnections) than giant polygons. In the Viking images we see that (1) old impact craters are increasingly buried with younger materials toward the basin center and that (2) a large number of superposed impact craters on polygonal terrain exhibit distinctive ejecta morphologies such as rampart ejecta blankets and single and double lobate ejecta blankets. MOLA and MOC observations are consistent with the formation of polygonal terrain in the area of a former standing body of water, but polygon size seems too large to be readily accounted for by desiccation or freezing processes alone. On the basis of the observed characteristics we propose that the giant polygons are primarily of tectonic origin, being caused by uplift of the floor of the Utopia basin. There are several candidates for the cause of uplift: (1) the removal of the load representing former standing bodies of water. Water thickness estimated for the Utopia basin is about 1 km. In this scenario, removal of this load(comparable to loads placed on terrestrial continental lithosphere by the Laurentide and Scandinavian ice sheets) by loss of the water or ice could cause uplift of the floor and polygon formation. (2) The freezing and expansion of residual water buried by sediments or in the near subsurface, or (3) a combination of these. From the discussion of alternative models in the light of the MOLA and MOC data we conclude that formation of giant polygons by tectonic uplift of the basin floor is most likely and deserves further investigation. ¿ 2000 American Geophysical Union