CpG islands (CGIs) are clusters of CpG dinucleotides in GC-rich regions and represent an important gene feature of mammalian genomes. Several algorithms have been developed to identify CGIs. Here we applied Support Vector Machine (SVM), a machine learning approach, to classify CGIs that are associated with the promoter regions of genes. We demonstrated that our SVM-based algorithm had much higher sensitivity and specificity in classifying promoter-associated CGIs than other algorithms, and had high reliability. The advantages of SVM in our method and future improvements were discussed.