Recent advances in genomic and proteomic technologies have revolutionized our way of classifying cancers. These high-throughput technologies allow the use of powerful and multivariate bioinformatic approaches to develop molecular classifiers. These classifiers can then be used to distinguish different types of tumors based on their molecular profiles. This is particularly important for heterogeneous groups of tumors such as sarcomas. Although sarcomas have a variety of histological appearances, the distinction among some of the diagnostic groups is vague. Therefore, molecular classification provides a new way to distinguish histologically similar but molecularly different types of sarcomas, and hence improves tumor diagnosis and stratification. In addition, identification of discriminatory genes that carry information to differentiate clinical subtypes of sarcomas will further elucidate the underlying molecular pathways and pathological mechanisms of these tumors. In this article, we review some current methods used in genomic and proteomic profiling, outline the approach of using bioinformatic techniques to develop a molecular classifier, and discuss some recent examples to illustrate the use of molecular classification to distinguish different types of sarcomas and understand the biology of these tumors.