The binding states of cysteine residues on protein carry important information about the function and structure of proteins. In this research, support vector machine (SVM) was used for the prediction of bonding states of cysteines. We found that the oxidation states of amino terminus cysteines infer the oxidation states of other cysteines in the same protein chain. By including the oxidations states of up to two cysteines on the amino terminus and the flanking sequences around the cysteine as well as the global protein information, accuracy of 95% is achieved for the data set consisting of 4136 cysteine-containing segments extracted from 969 nonhomologous proteins.