This paper examined the petroleum futures volatility comovements and spillovers for crude oil, gasoline, heat oil and natural gas. The results of volatility analysis were used to calculate the optimal two-petroleum portfolio weights and hedging ratios. The data used in this study was the daily data from 2009 to 2014. The three Multivariate GARCH models, namely the VAR (1)-diagonal VECH, the VAR (1)-diagonal BEKK and the VAR (1)-CCC, were employed. The empirical results overall showed that the estimates of the multivariate GARCH parameters were statistically significant in almost all cases except in the case of RGASOLINE with RNG. This indicates that the short run persistence of shocks on the dynamic conditional correlations was greatest for RCRUDE with RHEATOIL, while the largest long run persistence of shocks to the conditional correlations for RCRUDE with RGASOLINE. Finally, the results from these optimal portfolio weights base on the VAR (1)-diagonal VECH estimates suggested that investors should had more heat oil than crude oil and other petroleum in their portfolio to minimize risk without lowering the expected return.