Over the past two decades models have been developed which suggest that the Asian continent has been formed since the Late Palaeozoic by the accretion of continental blocks derived from the northern margin of Gondwana. Sumatra, forming the southwestern margin of the Southeast Asian promontory (Sundaland), is considered to be composed of fragments of continental plates and magmatic arcs which were derived from Gondwana during the Late Palaeozoic and Mesozoic. The Indochina Block forms the core of Sundaland, extending into the eastern part of the Malay Peninsula. The greater part of Sumatra is considered to form part of the Sibumasu Block which accreted to the Indochina Block along the Bentong-Raub Suture in the Triassic. A model has been proposed in which the southern part of the Sibumasu Block in the western part of the Malay Peninsula, and Sumatra, is divided into Malacca and Mergui Microplates by the Mutus Assemblage which represents another Triassic suture. A review of the Permo-Triassic stratigraphy of Malaya and Sumatra provides no support for this model. Comparison of the Permo-Carboniferous stratigraphy and palaeontology of northern Sumatra with that of the Malay Peninsula and Peninsular Thailand, and in particular the occurrence of tilloids, links Sumatra firmly to the rest of the Sibumasu Block to the north. Comparison with the Permo-Carboniferous stratigraphy of Bonaparte Gulf region of northwest Australia shows a mirror image relationship, suggesting that the Sibumasu Block separated from this part of the Gondwanan margin in the mid-Permian. On the other hand Permo-Carboniferous rocks in Central Sumatra contain a Cathaysian fauna and flora, which relates this area to the Indochina Block rather than to Sibumasu. This anomaly was recognised early in the study of the geology of Sumatra and led to the proposal of the Djambi Nappe, thrust over Sumatra from the east. The Cathaysian fauna and flora is associated with an Early Permian Volcanic arc. It has been suggested that this was an independent island arc accreted to the western margin of Sibumasu, but from the relationships of the volcanic rocks to Permian sediments and the underlying basement, it is most probable that this arc was developed on the margin of the Cathaysian Block and was emplaced in its present position outboard of Sibumasu by strike-slip faulting. The most recently accreted pre-Tertiary tectonic unit on Sumatra is the Woyla Group, a Jurassic-Early Cretaceous oceanic volcanic arc which, together with its associated accretionary complex of oceanic crustal material, was thrust over the western margin of Sumatra in the mid-Cretaceous. Earlier plate models for the development of Sumatra are reviewed, and a revised model is proposed. However there are still many difficulties in interpreting the stratigraphy and the tectonic development of Sumatra which will require further detailed study, aimed at resolving the many outstanding problems.
