Improved methods of collection have led to the discovery of much better amphibian fossils from localities in the Upper Triassic Maleri formation of the Pranhita Godavari Valley. Some preliminary observations regarding the Maleri sediments and their relationship with the overlying and underlying rocks are noted. A composite restoration of the skull is given. The pattern of the dermal roofing bones, the shape of the skull and the position of the orbits clearly indicate that the material belongs to a metoposaurid amphibian. A new metoposaur species, Metoposaurus maleriensis is proposed and a description is given of its skull, braincase and palate. The presence of a prominent additional nasal bone on each side, exposed on the skull roof in one of the skulls of M. maleriensis and its absence in other skulls of this metoposaur is noted and the possible explanation for this condition is put forward. The position of the orbits varies according to the size of the skull; a comparison is made between the two more complete specimens to show that the orbits are more posteriorly placed in larger skulls. It is suggested that this difference in orbital position is probably due to the relative rate of activity of two zones of intensive growth in the skull, anterior and posterior to the orbits. Postcranial material includes three clavicles, interclavicle, humerus, ischium and eight intercentra including that of an atlas. These have also been assigned to M. maleriensis on the basis of close association and osteological characters. Study of the Maleri metoposaur and its comparison with the other adequately known genera, Metoposaurus from the Keuper of Europe and 'Eupelor' from the Upper Triassic of North America, has demonstrated the weak foundation on which metoposaur genera are based. The position of the lacrimal bone has been used as one of the main morphological characters distinguishing between the previously recognized metoposaur genera. However, this character is variable within a subspecies recognized by Colbert & Imbrie, which suggests caution in its use in distinguishing between genera. The recognition of this feature further emphasizes the extremely close similarity between the metoposaur genera, already commented on by previous authors. The suggestion is made that these genera are morphologically indistinguishable and a revised taxonomy of the metoposaurs is put forward. The upper and lower stratigraphic limits of the metoposaurs are examined, and it is found that the metoposaurs are restricted to horizons equivalent to Carnian and Norian. The fauna associated with the metoposaurs in different parts of the world is discussed, and it is concluded that the age of the Maleri fauna is not younger than Middle Norian and probably not older than Carnian.