This paper tests empiricaly the Balassa-Samuelson (BS) hypothesis using annual data for Australia. We applied the ARDL cointegration technique developed by Perasan et al. (2001) and found evidence of a significant long-run relationship between real exchange rate and Australia-US productivity differential during the period of 1950-2003. We found that a one per cent increase in labour productivity in Australia relative to the US will lead to 5.6 per cent appreciation in the real exchange rate of Australia. We suspect that the elasticity coefficient is "ever-estimated" due to the exclusion of relevant explanatory variables. The dynamics and the determinants of the real exchange rate movements are numerous; they include terms of trade, interest rate differentials, net foreign liabilities among others along with labour productivity differential.