Poor ground cover and low soil moisture are necessary conditions for the occurrence of dust-related phenomena, and rainfall is the best single measure of water availability which controls both of these conditions. To improve our understanding of the role of the rainfall factor in dust activity, rainfall in the dust source area was compared with the frequency of occurrence of dust events at Mildura, Australia, for the period from 1960 to 1989. Dust activity was most intensive during the summer months and prevailing wind directions suggested that the dust came mainly from the west and south west. Major dust episodes were generally preceded by dry spells lasting through the previous autumn and winter seasons. The annual dust day total was found to correlate best with the previous year's rainfall, suggesting a high sensitivity of dust activity to the rainfall regime, probably because of the response of annual crops and pastures in the source areas to autumn and winter rainfall and their effect on soil cover. A simple model was developed to predict the frequency of summer dust event days based on rainfall during the previous autumn. It performed reasonably well with an average absolute error of three dust days.