Variations in crustal thickness from the Great Plains of Kansas, across the Colorado Rocky Mountains, and into the eastern Colorado Plateau are determined by receiver function analysis of broadband teleseismic P waveforms recorded during the 1992 Rocky Mountain Front Program for Array Seismic Studies of the Continental Lithosphere (PASSCAL) experiment. The receiver functions are calculated using a time domain deconvolution approach and are interpreted in terms of a single crystal layer, with thickness determined by a grid-search comparison of observed receiver functions with synthetics. The average crustal thicknesses determined by these methods are Kansas Great Plains, 43.8¿0.4 km; Colorado Great Plains, 49.9+1.2 km; Colorado Rocky Mountains, 50.1¿1.3 km; and northeast Colorado Plateau, 43.1¿0.9 at latitudes of 38¿--40¿N. The main variations in crustal thickness that we observe are between the Kansas Great Plains and the Colorado Great Plains and between the Rocky Mountains and the Colorado Plateau. There is not a significant crustal thickness difference between the Colorado Great Plains and the Colorado Rocky Mountains. Together with gravity data and mass balance calculations, these results are incompatible with the hypothesis that the compensation of the Rocky Mountains relative to the Great Plains is accommodated purely by an Airy-type crustal root or any other mechanism that restricts compensation solely to the crust and requires significant support for the excess topography of the Rocky Mountains to come from the mantle.

Models with a rigid elastic plate may match receiver function estimates of crustal thickness but underpredict the amplitude of the gravity low over the Rockies. Our favored model includes lateral variations in crustal velocities obtained from refraction studies and crustal thickness variations constrained by the receiver functions. These models indicate that there is a profound transition in mantle density structure near the eastern range front. ¿ American Geophysical Union 1995.