The current views on the relation between latero-sensory lines and dermal bones are briefly reviewed, and a modified version is proposed. Published photographs of a number of parietal shields of crossopterygians other than coelacanths have been re-photographed, and enlarged outlines brought to a common frame of reference and compared. On this basis the significance of variations in Osteolepis that have two parietal bones on one side and the usual single parietal on the other is considered, the parietal shields of some genera of osteolepiforms, and also the equivalent area in ichthyostegids, are compared with one another, and the differences between the parietal shields ofosteolepiforms, porolepiforms and rhizodontiforms are discussed. These last are believed to be due to the infraorbital lateral line having come to be situated at different relative locations at the beginning of skeletogenesis; the latero-sensory ossicles that it formed consequently came to act as foci for the development of different combinations of bones. Published work on the skull-roof of dipnoans is also considered, with particular reference to the question of how bones I and B became latero-sensory. The variations in dipnoan pit-lines seem to have been similar in principle to those of the superficial sensory lines in placoderms. A scheme relating the dermal skull-roofs of the main groups of gnathostomatous fishes is proposed.