Data from rock-dredging have often been used to infer that oceanic fracture zones provide a 'window' into layers of the oceanic crust lying at a depth below the surface that is approximately equivalent to the vertical offset of the fracture zone, and thus permit the reconstruction of a crustal stratigraphy for the whole of acoustic layer 2 (commonly considered to have an average thickness of ~2 km) and, in some interpretations, for the upper part of layer 3. Alternatively, it has been suggested that fracture zones are preferential sites of serpentinite mega-dykes differing in composition from layer 3 but containing inclusions of the third layer. The published data indicate that basalts and basaltic rubble are abundant in fracture zones and, on analysis, do not justify the assumptions that have been made. The structure of fracture zones limits the possible extent of crustal sections exposed on their walls. Moreover, it is suggested that rocks of different layers of the lithosphere can be emplaced in the transform domain due to the dynamic of the transform fault system, itself.