It is generally accepted that transforming growth factor beta (TGFbeta) is both a tumor suppressor and tumor promoter. Whereas loss or attenuation of TGFbeta signal transduction is permissive for transformation, introduction of dominant-negative TGFbeta receptors into metastatic breast cancer cells has been shown to inhibit epithelial-to-mesenchymal transition, motility, invasiveness, survival, and metastases. In addition, there is evidence that excess production and/or activation of TGFbeta by cancer cells can contribute to tumor progression by paracrine mechanisms involving neoangiogenesis, production of stroma and proteases, and subversion of immune surveillance mechanisms in tumor hosts. These data provide a rationale in favor of blockade of autocrine/paracrine TGFbeta signaling in human mammary tumors with therapeutic intent. Several treatment approaches are currently in early clinical development and have been the focus of our laboratory. These include (1) ligand antibodies or receptor-containing fusion proteins aimed at blocking ligand binding to cognate receptors and (2) small-molecule inhibitors of the type I TGFbeta receptor serine/threonine kinase. Many questions remain about the viability of anti-TGFbeta treatment strategies, the best molecular approach (or combinations) for inhibition of TGFbeta function in vivo, the biochemical surrogate markers of tumor response, the molecular profiles in tumors for selection into clinical trials, and potential toxicities, among others.