This paper studies the occurrence and causes of the regime-switching of two representative commod-ity indexes，S＆P GSCI ( Standard ＆ Poor’s Goldman Sachs Commodity Index) and DJCI ( Dow Jones Com-modity Index) ，by using the modified Markov Ｒegime-Switching model with the dummy variables that reflect the changes of regimes both before and after the financial crisis． The results show that the suggested model performances better in characterizing the periodic variation of return volatility，which possesses a feature of “medium-high-medium high-low-medium”． Around the turning point，the considered commodity index is shaped as a “V”type． It also reveals that liquidity shocks in the process of financial crisis could alter the probability of regime-switching． The empirical findings could provide some references，to some extent，for the prediction of the trend of commodity markets and the emergence of the stage turning points．