There are a number of performance measures that are widely available to aid superannuation savers and fund managers for predicting future funds' performance. The three traditional measures of funds' performance that have been widely established in the literature are the Treynor, Sharpe and Jensen measures. However, these measures are also associated with a number of shortcomings that have been highlighted in the literature. This thesis seeks to address some of their limitations by employing an increasingly popular benchmarking tool in productivity research called data envelopment analysis (DEA). This study shows the benefits of the proposed approach and applies the DEA technique to measure the relative efficiency of the Australian multi-sector superannuation investments, namely aggressive, growth, balanced, moderate and defensive funds. The DEA analysis shows that the primary source of funds' inefficiencies within the investment classes lies in scale inefficiency. The results show that the aggressive funds perform more efficiently than the funds in other investment classes. Also, this thesis finds a significant relationship between poorer performance and higher fund fees. However, this relationship does not hold for the aggressive funds, indicating that they may have superior market information.