Erosion beveled the Laramide Front Range uplift in Colorado to a surface of low relief by the end of the Eocene. This study uses J. A. Wolfe's new multivariate climate analysis techniques to determine the paleoelevation of this regional surface by examining the overlying 35 Ma Florissant flora. A multiple regression model explaining 93.3% of the variance in mean annual temperature was developed using Wolfe's dataset of 31 leaf physiognomic character states for 86 modern vegetation sites. These character states were scored on 29 species collected from one facies of the Florissant Lake Beds. The paleotemperature estimate of mean annual temperature (10.7 ±1.5 °C) derived from these data, when combined with sea-level temperature and terrestrial lapse rate, implies a late Eocene paleoelevation of 2.4-2.7 km. Pliocene uplift is thus not required to explain the present elevation of 2.5 km. It is unclear when and why the southern Rocky Mountains achieved this elevation. Magmatic crustal thickening can explain the late Eocene high elevation of the southern Rockies, but neither this mechanism nor compressive thickening explains why the Great Plains, which are tied to the Florissant elevation by the Wall Mountain Tuff, were also high. This paleoelevation estimate indicates that regional surfaces of planation could be formed at high elevation in the Eocene, probably because of peculiarities of the Eocene climate.