Microscopic charcoal fragments preserved in lake and swamp sediments with pollen and other microfossils provide evidence for past burning regimes. Some problems with the interpretation of charcoal records are similar to those of pollen analysis, but other factors must be taken into account. Among these are: (a) fires are discrete, irregular events, highly dependent on weather and vegetation; (b) most charcoal is transported by water, rather than in smoke, so the amount of charcoal deposited will depend on rainfall after the fire, as well as on the amount produced; and (c) more charcoal in sediments may mean more or less frequent fires, depending on the rate and amount of fuel build-up in different vegetation types and the resolution of sediment sampling. Reconstructions of vegetation and fire histories in Australia suggest that climate has been the major determinant of vegetation and, on the broad scale, Aboriginal burning has had little effect.