Soft tissue sarcomas represent a heterogeneous group of tumors and include over 50 histotypes. Some of these tumor types are characterized by specific chromosomal translocations, whereas other types show complex genetic aberrations. The recent developments within gene expression technologies have now been applied to studies of soft tissue sarcomas (STS) and the first results indicate that genetic signatures are useful for classification and diagnosis. Distinctive expression profiles have been found in e.g. gastrointestinal stromal tumors (GISTs), synovial sarcomas, malignant peripheral nerve sheath tumors (MPNSTs), and in subsets of liposarcomas. The more pleomorphic tumor types, such as high-grade variants of leiomyosarcomas, malignant fibrous histiocytomas (MFHs), fibrosarcomas, and subtypes of liposarcomas, show a greater variability among the expression profiles, but interestingly subsets with distinctive expression profiles can be identified also among these tumors. The data available place many of the genes hypothesized to be involved in the development of a certain type of STS, such as the KIT gene in GIST development, among the top discriminating genes. Thereby expression profiling provides novel insights into the pathogenesis of STS. Although much work remains to be done to validate the data and to define optimal discriminating gene lists, the current lessons from gene expression studies in STS are encouraging and imply that genetic signatures may serve as diagnostic and prognostic markers and may help identify novel therapeutic strategies.