The true role of transforming growth factor beta1 (TGFbeta) on angiogenesis is in question. Several in vitro studies have shown inhibition of proliferation and migration of endothelial cells (EC). However, some investigators have observed that TGFbeta stimulates the formation of EC tubes in vitro. Fewer in vivo studies have been performed, but these also show discrepancies: some report angiogenic induction and at least one reports inhibition. We used the disc angiogenesis system (DAS) to measure the in vivo effect of TGFbeta. Discs containing 1 ng to 2000 ng of TGFbeta were placed subcutaneously in mice, removed after a growth period of 14 days, measured by three different techniques, and compared with spontaneous growth controls and with positive controls containing prostaglandin El. Tritiated thymidine was used to determine proliferation of EC. The discs were also examined morphologically for patterns of vessel and stromal proliferation. The contribution of native TGFbeta to the spontaneous angiogenesis of wound healing was tested using a monoclonal anti-TGFbeta antibody. The combined effect of TGFbeta and fibroblast growth factor (bFGF) was studied by using suboptimal doses of both (500 ng and 10 microg, respectively). Although TGFbeta doses of 1 ng to 500 ng failed to induce angiogenesis, 1000 ng induced a significant level of angiogenesis which was maintained at 2000 ng. This effect was the same regardless of the method of quantification: centripetal growth of the vessels, size of fibrovascular growth area, or amount of incorporation of tritiated thymidine. The anti-TGFbeta antibody decreased the spontaneous vascular growth below the level of controls containing irrelevant IgG. The combination of TGFbeta and bFGF at suboptimal doses did not increase or decrease the angiogenic response. Discs containing TGFbeta showed more collagen and greater accumulation of neutrophils than control discs or discs containing other cytokines. In conclusion, TGFbeta1 is angiogenic in vivo, when it reaches a threshold of 1 microg, but is not angiogenic at doses of 1 to 500 ng. Endogenous TGFbeta contributes to spontaneous (wound healing) angiogenesis. At the suboptimal doses of TGFbeta and bFGF used, there is no evidence of a combined angiogenic effect. The angiogenic effect of TGFbeta is probably indirect, requiring recruitment of leukocytes that secondarily release angiogenic substances. This secondary effect may explain some of the discrepancies between the in vitro and in vivo effects of TGFbeta.