The root mean square velocities of the major continents with respect to the hot spots were estimated for the past 180 m.y. by combining previously published plate reconstructions. These velocities agree well with velocities relative to the spin axis inferred from palemagnetic data. From ~169 m.y. B.P. to ~144 m.y. B.P., the hot spot data indicate that Laurasia moved rapidly, ~70 mm/yr, but the reconstructions permit a velocity as low as 50 mm/yr. We also estimate that from ~63 m.y. B.P. to ~48 m.y. BP., India moved rapidly (~140+30-20 mm/yr), faster than the present velocity of any plate and more than twice as fast as the present velocities of several subducting oceanic plates (those attached to downgoing slabs along a substantial fraction of their margin). The data require no other continent to have exceeded a velocity of 45 mm/yr during the last 180 m.y. In agreement with previous studies based on paleomagnetic data and in agreement with previous studies based on no-netorque analyses, we conclude that the continents have moved faster, occasionally much faster, than at present. From this we infer that continental lithosphere is unlikely to have much greater resistance to plate motion than oceanic lithosphere. Instead, oceanic plates tend to move more rapidly than continental plates because the subduction and rapid motion of a plate that includes no continents tends to persist longer than the subduction and rapid motion of a plate that includes one or more continents. On the basis of the very rapid early Tertiary motion of India, we conclude that the velocities of suducting plates can vary by a factor of 2. Therefore models of the driving forces of the plates that predict a uniform velocity of subducting slabs are invalid.