Assessment of current Proterozoic palaeomagnetic data and the assumptions necessary for applying them to tectonic problems shows that first order tectonic phenomena (involving the creation and/or destruction of Atlantic-sized oceans) should now be detectable. Although lesser relative motions are difficult to prove or disprove in individual cases, it is shown by reference to Africa (from where new data for the Palabora Igneous Complex and lavas of the Transvaal System are presented which contribute to this discussion) that their occurrence between some of the African cratons is improbable. It is necessary, then, to entertain the notion that some tectonically deformed belts, though not necessarily all, developed without associated destruction of large amounts of oceanic crust on their site. Some ways in which palaeomagnetic study may throw further light on this problem are suggested, but a wide range of geological techniques will be needed to solve it, and to assess its broader implications.