THE Alpine mountain chain is generally accepted to be the product of continent-continent collisions. In this belt the zone of recent tectonic activity is wide (up to 2,000 km in Tibet) and crustal thickness in places is 1.5-2 times the average continental crust, presumably due to the inability of light continental material to sink into the astheosphere. Under the Himalayas, for example, the crust is 70 km thick1. Furthermore, as indicated by seismicity, the active collision zone here includes not only the highly deformed Himalaya belt but also the entire Tibet plateau. Major wide mountain belts exist, morphologically similar to the Alpine belt, in regions which do not experience continental collision, such as western North America, Alaska, east Siberia and the Andes. The crustal thickness here can also be very great, up to 70 km in the Andes2. All are seismically active, wide, highly deformed and include high plateaus of various sizes. Many of these wide erogenic belts also exhibit great geological complexities which are not simply explained by the model of an oceanic lithosphere under-thrusting a continental lithosphere. We suggest, therefore, that the circum Pacific mountain belts may be the result of past continental collisions, similar to those associated with the Alpine belt. We summarise the evidence for the incorporation of past continental masses around the Pacific Ocean. Holmes3 has given a compelling case for large continental land masses during parts of late Palaeozoic to early Tertiary to the west of North America such as Cascadia and Llanoria4. The land includes conglomerates derived from crystalline sialic rocks which have since disappeared. Hamilton5 and Davis and Armstrong6 suggested that the Klamaths were originally some distance offshore to the west and that the Permo Triassic Sonoma Orogeny results from an arc continent collision.