Fossils provide an independent criterion for testing structural hypotheses generated from field evidence or geochemistry, as well as functioning in their traditional role in the dating of rocks or determining biogeographic realms. On a global scale, the identification of exterior, oceanic biofacies may be of use in discriminating the course of former continental margins. The isograptid graptolite biofacies (Early to Middle Ordovician) can be distinguished from contemporary epicratonic graptolite biofacies; its worldwide occurrence corresponds with continent margins reconstructed from other criteria. Hence a new isograptid record from Irian Jaya (E New Guinea), although based on a small sample, suggests that there may have been an Ordovician oceanic site there. On a continental scale, the tectonic juxtaposition of different biofacies, or of different stratigraphical packages, contributes to an understanding of tectonics and can serve to quantify displacement: the accretionary prism hypothesis for the Southern Uplands of Scotland is supported and contrasted with the rearrangement of biofacies in the Cambro–Ordovician Cow Head Group of western Newfoundland which reveals not only the original shelf-margin–slope topography (now vanished), but also provides an estimate of the movement involved in the emplacement of one Taconic allochthon. On a more local scale, faunal evidence is used in assessing the original position of 'suspect' marginal areas, using the simplest possible model. This shows, for example, that Anglesey was not greatly separated from Gondwana in the early Ordovician—but strike-slip movement cannot be disproved. Local changes in relative sea-level are determinable from sediments and faunas; in marginal belts these may differ from the global eustatic curve, and when this happens it is a measure of local tectonic activity. Particular sea-level curves may serve to define structurally coherent units within mobile belts. New sea-level curves are contrasted for a segment of N Wales (north of the Bala fault) with that for central SW Wales (Arenig–Llandovery); the S Welsh area mostly responded passively to the eustatic curve, which can adequately account for shifts in facies and faunas, while N Wales had a curve often at odds with eustasy and reflecting its structural history.