This article investigates the potential excess returns from a value investment strategy in Australia and uses the cross-section methodology of Fama and MacBeth (1973) and the time-series approach of Fama and French (1993) to examine the role of size of the firm and the book-to-price ratio in explaining average stock returns. The first finding from the study is that over the period 1992 to 2001, value portfolios outperformed glamour portfolios annually by an average of 11.2% with portfolios sorted on the book-to-price (B/P) ratio, 8.2% with portfolios sorted on the cash flow-to-price (C/P) ratio, and 8.2% for portfolios sorted on the earnings-to-price (E/P) ratio. We also find that the FamaFrench three-factor model, based on size, book-to-price and the market risk premium, explains average returns better than a one-factor model (i.e. CAPM) based on the market risk premium alone. Consistent with the previous Australian study by Halliwell et al (1999) over 1981 to 1991, the 5MB (small size minus big size) factor is significant across most portfolios, but the HML (high B/P minus low B/P) factor is significant only in portfolios with the largest market capitalisations. The results in Australia differ from Fama and French (1993), who found the HML factor to be significant across almost all 25 size and B/P quintile portfolios in the U.S. We conjecture that different market characteristics in Australia, in particular the large percentage of very small market capitalisation firms, may explain the dissimilar finding on the HML factor in Australia.