Mantle xenoliths from western Saudi Arabia, some 220 km east of the Red Sea margin, indicate temperatures as high as 900 ¿C in the mantle at about 40 km depth beneath an area of the Arabian Shield where surface heat flow is lower than the world average. This suggests that there has been a large increase in the mantle temperature that has not had time to equilibrate at the surface. The amount of uplift and the difference between the observed xenolith temperature and that obtained by projecting the surface heat flow to depth suggests that the lithosphere has been thinned by 30 to 100 km to its present calculated thickness of about 80 km. Calculations of the time necessary for thermal reequilibration indicate that the lithospheric thinning probably occurred less than 20 m.y. ago, so its postdates rifting which started about 30 Ma. If there was a prolonged upwelling of deep mantle material, resulting in a thinner lithosphere prior to rifting, there should have been enough time for some reequilibration at the surface, and the observed heat flow should be higher over the broad area of uplift adjacent to the Red Sea. Our calculations are consistent with a postrift age of uplift adjacent to the Red Sea as determined from regional geologic evidence and fission track ages on apatites from shield rocks. The Red Sea appears to have begun with alkaline volcanism on the Arabian Shield, followed by rifting. Uplift and lithospheric thinning adjacent to the Red Sea rift began 5--10 m.y. after rifting. This suggest that the Red Sea is a passive rift that formed because of two-dimensional plate motions, rather than as a consequence of mantle upwelling. A possible cause for the stress in the plates might have been the change of plate motions in the Indian Ocean that resulted from the collision of India with Eurasia. ¿ American Geophysical Union 1989