Mars may have substantially changed its average axial tilt over geologic time due to the waxing and waning of water ice caps through the phenomenon of climate friction (also called obliquity-oblateness feedback). Depending upon Mars' climate and internal structure, water caps of the order of 1017-1018 kg cycling with the obliquity oscillations could have either increased or decreased the average obliquity by possibly tens of degrees. This is in contrast to previous results, which indicated that 1017 kg carbon dioxide caps only increased the axial tilt. Since the south polar cap appears to be mostly uncompensated, Mars may be largely rigid on the obliquity timescale. Further, Mars may be a water-rich planet so that there is a large phase angle between insolation forcing and the size of the obliquity-driven water caps. A stiff, water-rich planet indicates the obliquity may have decreased over the eons. Such a decrease might account for the apparent youthfulness of the polar layered terrain, the idea being that fewer volatiles were available to be cycled into and out of the terrain at high obliquity because of more even insolation between equator and pole, so that the movement of volatiles produced thin layers or perhaps no layers at all. As the obliquity decreased, the insolation contrast between high and low latitiudes increased, and more volatiles might have shuttled into and out of the polar regions, forming the observed thick layers. In another but perhaps less likely scenario, Mars' average obliquity may have either increased or decreased until it became stuck at its present value of ~24¿. In this case the idea is that Mars' climate dynamics altered as the average tilt changed. Once the rate of increase in tilt caused by the deformation of the solid planet equaled the rate of decrease caused by the caps, the obliquity evolution ceased, leaving Mars at its present tilt. ¿ 1999 American Geophysical Union