Knowledge of the climatological processes that drive the Australian summer monsoon and the ecological consequences of this highly predictable seasonal cycle of wet summers and dry winters has greatly increased over the last several decades. Nonetheless, remarkably little is known about the antiquity of this climate system or how the geographic extent and duration of the monsoon have varied across temporal scales from the recent to the distant geological past. Late Pleistocene sedimentary records from lakes in the interior and north of the Australian continent have been used as a measure of the strength of the palaeo-monsoon. However, in the case of the inland lakes this approach is confounded because it is known that under the contemporary climate, full-lake conditions occur infrequently in response to high rainfall following either a strongly positive ENSO phenomenon (la Nina conditions) or following the continental incursion of tropical depressions.While the general absence of sedimentary and fossil records of the early Pleistocene and late Tertiary prevents elucidation of the timing of onset of the summer monsoon, the degree of biological adaptation of a diverse assemblage of organisms to the wet-dry tropical climate suggests that the summer monsoon climate is of great antiquity. This point is substantiated by a brief review of the ecophysiology of a range of Eucalyptus savanna plant species, the biophysics of the endemic 'magnetic' termite mounds and the reproductive cycle of a freshwater turtle. Assuming that the configuration of land and sea is of critical importance in controlling the strength of the Australian summer monsoon, then the notion of great antiquity of the monsoon is also consistent with the fact that for much of the Quaternary, and probably also during the late Tertiary, the relative extent of sea and land has been similar to that of today.The reliable, dry season, southeasterly trade winds create conditions favouring widespread, frequent bushfires. Under both European and Aboriginal management regimes most fires are started by people, although fires can be started by rain-free thunderstorms that occur during the transition between the dry and wet season when intense convection storms are common. Prior to human colonisation such storms would have been an important ignition source for landscape fires, and this may help explain why the north Australian biota is so remarkably tolerant of recurrent fires. The general absence of fossil evidence also frustrates attempts to determine what effect burning by late-Pleistocene human colonisation had on the north Australian landscapes. Given the tolerance of biota to fire, this change 262 Australian Geographical Studies © Institute of Australian Geographers 2002 may have only affected vegetation structure. Therefore the effect may be very difficult to detect from the pollen record once suitable longer-term records become available.The isotopic signature in extinct and extant ratite eggshells from late Pleistocene sediments in the Lake Eyre basin has been ...