A previous method for determining paleoaltitude from fossil floras has used the difference between sea level and upland paleotemperatures, which is then multiplied by the reciprocal of a ''normal'' lapse rate. The present study examines other variables that can significantly influence such results. Terrestrial lapse rates have been calculated for 39 areas of the world based upon mean annual temperature (MAT), mean warm month temperature (WMT), and mean cold month temperature (CMT). These show regional and seasonal variability that is explained by factors including continentality, elevation of large massifs or high base levels, cold air drainage, fog, mountain barrier effects, and the location of the subtropical high-pressure zone. Lapse rates and projected sea level temperatures based on data from the western United States show distinct regional patterns. These indicate that, within this region, the highest lapse rate and temperature values are correlated with (1) continental climates in the subtropical high-pressure zone, and (2) areas having relief where the lowest elevation land surfaces (base levels) occur at high altitude. The data are applied to evaluate and revise the methodology for estimating paleoaltitudes. Variables of climatic change, sea level fluctuation, and paleogeography are also considered. Limitations in application of the method are emphasized. The method is exemplified by application to the latest Eocene Florissant flora of Colorado, U.S.A. Although the results are conjectural, they strongly suggest that the late Eocene erosion surface in the southern Rocky Mountains occurred at a much higher paleoaltitude than previously thought. This high elevation may have resulted from earlier Laramide crustal thickening.
