This paper investigates the usefulness of commodities in investors' portfolios within a mean- variance optimization framework. The analysis di_x000B_ers from previous research by considering multiple investment tools including individual commodity futures contracts, three generations of commodity indices and by controlling for estimation error in portfolio optimization pro- cess. Rather generally, the results demonstrate that including individual commodities or the first- and second-generation commodity indices do little to enhance portfolio performance. Similarly, when an initial portfolio is diversi_x000C_ed, the risk-reducing ability of agricultural commodities is much weaker than identi_x000C_ed by previous research. In contrast, including the third-generation indices substantially improves the portfolio's Sharpe ratio by generating higher returns and lower risk.