In conclusion, we have demonstrated that a single growth factor, TGF-beta, can act in either an autocrine or paracrine mode to bring about, either directly or indirectly, all of the complex events which together lead to the formation of granulation tissue and tumor stroma. All of the participating cell types, lymphocytes, macrophages, endothelial cells, and fibroblasts have receptors for TGF-beta and many of them secrete TGF-beta as well. Although other growth factors such as platelet-derived growth factor and fibroblast growth factor clearly also participate in these processes, we have shown that TGF-beta, alone, is sufficient to initiate the cascade of events, probably through its ability to chemoattract and to activate macrophages and fibroblasts. Without question, neovascularization and appropriately timed matrix synthesis and degradation are central to embryogenesis as well. Fibronectin, in particular, has been shown to promote cell adhesion and cell migration throughout embryogenesis (Hynes and Yamada, 1982; Rovasio et al., 1983). We propose that TGF-beta will be found to be an important mediator of embryonic development, not only by control of angiogenesis and matrix synthesis, but also by exerting direct local effects on cellular growth and differentiation (reviewed in Roberts and Sporn, 1987).