Postglacial rebound appears to have been observed gravitationally by the Lageos satellite. Sixty‐four observations of the orbital node made over a 5‐year time interval reveal an acceleration of ( −8.1±1.8)×10−8 arc sec d−2 due to a source which is not presently modeled in the GEODYN orbit determination computer program. This acceleration cannot be explained by the ocean tide with an 18.6‐year period, assuming it to be an equilibrium tide. Instead it seems to be due to postglacial rebound, which changes the J2 coefficient in the spherical harmonic expansion of the earth's gravitational field at the rate of ( −8.2±1.8)×10−19 s−1 this, in turn, accelerates the node. This rate does not agree with the −32×10−19 s−1 predicted by Wu and Peltier's L2 model, which has upper and lower mantle effective viscosities of 1021 and 1022 Pa s, respectively. It does agree well with their L1 model, which gives about −10×10−19 s−1. Since the effective viscosity is 1021 Pa s throughout the entire mantle in the L1 model, the results support the contentions that (1) the effective viscosity is near 1021 Pa s everywhere in the mantle, and (2) this relatively low value for the effective viscosity may have permitted several degrees of polar wander due to glaciation during the Quaternary Ice Age.