DETAILED analyses of deep-sea sediment cores have produced spectacular advances in the knowledge of Quaternary stratigraphy and palaeoclimates. These have allowed the rigorous testing of hypotheses put forward to explain climatic fluctuations1 and the construction of global atmospheric circulation models for dated periods in the past2. Unfortunately, these studies do not provide evidence of the important climatic variable, precipitation, changes in which could be critical to the understanding of glacial advance and retreat and to atmospheric circulation patterns generally. Only sequences from terrestrial sediments can provide the necessary information on precipitation. Here again the majority of dry land sequences are of limited value (with certain exceptions3,4) because they predominantly reflect temperature fluctuations or, where changes in effective precipitation are evident, sequences tend to be short or discontinuous. Here I outline and assess the significance of this sequence from tropical Australia which provides a long continuous precipitation record.