It has been argued that there are certain idiosyncrasies in Australian petrol price behaviour. To the extent that these idiosyncrasies result in large magnitude differences in petrol prices, they may be exploited by consumers to significantly reduce their household expenditure on the product. Similarly, such seasonalities may influence retailers in their purchase and storage decision. The objective of this paper is to test for seasonalities in the Australian retail petrol market. The approach adopted is similar to that for determining calendar anomalies as documented in the financial and commodity markets literature. We find that a monthly seasonal effect is pronounced with petrol prices lower in the months of February-May and highest in July and August. A day-of-the-week effect is also apparent and is manifest in all petrol prices for capital cities (Adelaide, Brisbane, Melbourne and Sydney) across various years. However, the half-month effect, as is common in stock returns, is not observed. Moreover, contrary to popular belief that petrol prices are higher surrounding holidays, no evidence of the holiday effect is found. In Brisbane and Melbourne, petrol prices also have some relationship to the mood of consumers, as proxied using weather conditions. This is not observed in Adelaide and Sydney. (C) 2000 Elsevier Science Ltd. All rights reserved.