In this study, productivity growth in 35 Australian universities is investigated using non-parametric frontier techniques over the period 1998-2003. The five inputs included in the analysis are full-time equivalent academic and non-academic staff, non-labor expenditure and undergraduate and postgraduate student load while the six outputs are undergraduate, postgraduate and Ph.D. completions, national competitive and industry grants and publications. Using Malmquist indices, productivity growth is decomposed into technical efficiency and technological change. The results indicate that annual productivity growth averaged 3.3% across all universities, with a range from -1.8% to 13.0%, and was largely attributable to technological progress. However, separate analyses of research-only and teaching-only productivity indicate that most of this gain was attributable to improvements in research-only productivity associated with pure technical and some scale efficiency improvements. While teaching-only productivity also contributed, the largest source of gain in that instance was technological progress offset by a slight fall in technical efficiency. (Contains 1 figure and 2 tables.)