This paper examines calendar effects in Australian daily stock returns from 6 January 1958 to 30 December 2005. Three calendar effects-day-of-the-week, turn-of-the-month and month-of-the-year-are examined using parametric tests and a regression-based approach. The results indicate that the Australian market is characterised by seasonality of all three forms, with Tuesday, September and the second trading day of the month the most significant. However, there is also evidence of parameter instability and structural breaks in these relationships, with day-of-the-week effects becoming less important in the post-1987 crash period.