Abundant paleomagnetic data exist for ophiolites, but they are not without problems. Structural corrections are sometimes difficult to ascertain, but more importantly, the age of the magnetization can be uncertain. For example, typical in-situ ocean-floor rocks have abundant titanomagnetites, whereas the magnetic minerals in ophiolites are usually Ti-poor. This suggests alteration at some stage after formation and before or during obduction. The magnetizations of ocean-island basalts, on the other hand, seem to survive the accretion/obduction process better than those of ophiolites, as will be illustrated with examples from the northern Appalachians. Still, well-known ophiolite complexes in the former Tethys (Oman, Xigaze, Troodos) have yielded important information about paleolatitudinal displacements. Collectively these data suggest that ophiolites do not appear to be typical ocean-floor basalts formed at a ridge near the middle of an ocean. Their paleolatitudes in the Tethys and Iapetus reflect origins near the continental margins, possibly all in back-arc basin settings. It can be concluded that typical mid-oceanic lithosphere seems to subduct without leaving much of a trace at the surface. Ocean-island basalts, in contrast, have a better chance of preservation near the surface. Whereas fully subducted oceanic lithosphere is no longer of use to paleomagnetists, it can nevertheless be used by paleogeographers in that its presence in the mantle can be recognized by tomographic imaging techniques. Examples will be briefly discussed of the paleogeographic inferences made from mantle P-wave velocity anomalies under Asia.