The broad trends of climatic evolution in the Southeast Indian Ocean region during the Tertiary have become clearer with the accumulation of data from Legs 28 and 29 of the Deep Sea Drilling Project. The same events which are reflected in the deep-sea sediments clearly influenced the major continental areas of Australia and Antarctica and the vegetation which covered those regions. In this review, data relating to sea-surface temperatures, to land and sea positions, and to the extent of the Antarctic ice-cap, are used to derive climatic models for the Palaeocene to latest Miocene interval: the models are correlated with vegetation history deciphered from palynological data. For most of the Palaeocene, Australia and Antarctica were joined, and lay in high southern latitudes. Sea-surface temperatures were relatively high, and there is no evidence for Antarctic ice. The postulated atmospheric circulation patterns suggest a zone of westerly winds confined to 60–80°S, and, north of that, a wide zone of erratic circulation, with deep inland penetration of rain-bearing winds. Palynological evidence from Australia indicates rainforest in the southeast, with an extension to presently arid inland South Australia. In the Eocene, Australia and Antarctica were separated by a wide gulf; water temperatures were high — in the vicinity of 20°C on the Campbell Plateau — and evidence for Antarctic ice development is meagre. Atmospheric circulation was probably sluggish, with patterns similar to those of the Palaeocene. Vegetational data show the spread of rainforest communities across Australia from the east to the southwest, and locally, inland to central Australia. Vegetation zonation appears to have been minimal on that continent. In Antarctica, late Eocene vegetation was poorly diversified. For the Oligocene, evidence for a pronounced temperature drop is both isotopic and sedimentological, and includes the first record of ice-rafting near Antarctica. Intensification of atmospheric circulation seems likely. The vegetation records suggest lowered floristic diversity in Australia, and the persistence of vegetation, again of low diversity, in Antarctica into the late Oligocene. For the Miocene, the marine record suggests the development of a major Antarctic ice-cap; in consequence, atmospheric circulations continued to intensify. Continued northward drift of Australia meant increasing aridity in regions to the north and northwest; the south and east were watered by rain-bearing westerly systems and by embryonic Trade Winds. Vegetation data in Australia suggest rainforest cover in the southeast and on the east coast, and, in the middle Miocene, the development of grasslands in interfluvial areas in the central part of the continent. The latest Miocene was marked by an intense and sudden chilling which is clearly documented in Southern Ocean sediments. This event must have caused marked precipitation decrease in much of Australia, and may coincide with the disappearance of pollen of the Nothofagus brassi type from regions west of the Great Dividing Range, although age control of this event is poor.