Solid tumors co-opt the body's endogenous extracellular proteolytic machinery for their invasion and metastasis. This is supported by a large number of independent observations ranging from histochemical and prognostic studies of cancer patient material to animal experiments. There are several extracellular proteolytic systems that are relevant in the context of cancer, but the plasminogen activation (PA) system and the matrix metalloproteases (MMPs) remain the most thoroughly investigated. Localization studies by immunohistochemistry and in situ mRNA hybridization in tumors of common human cancers have repeatedly identified members of the PA and MMP systems in stromal cells. The cancer cells, of epithelial origin, contribute PA and MMP components in some cases, but their contribution fades in comparison with the overwhelming expression of proteolytic components by fibroblasts, macrophages, endothelial cells, and other stromal cells. Ideal animal models of human cancers should recapitulate this fundamental proteolytic aspect of tumor biology. However, in the transplantable tumor models where PA or MMP components have been studied at the cellular level in vivo, this is most often not the case. Transgenic cancer models may provide a closer parallel to the human situation, in that PA and MMP components are synthesized by the tumor stroma. The pivotal role of stromal cells has been confirmed experimentally in mouse models in which the expression pattern of proteolytic components is strongly reminiscent of human tumors. In these models it is possible to reconstitute the wild-type tumor characteristics of proteolytically deficient tumor-bearing mice by transplantation with wild-type fibroblasts or hemapoietic cells. These studies collectively show that cancer-associated proteolysis is a collaborative effort of malignant cancer cells and various stromal cells--a collaboration in which stromal cells contribute the majority of the active proteolytic components that are necessary for the invasive behavior of the tumors. This cellular division of labor positions the stromal cells as prime targets for future research and possibly therapy. Vascular endothelial cells are already the focus of intense therapeutically relevant research, but tumor-associated fibroblasts, macrophages, neutrophils, lymphendothelial cells, etc. provide additional largely unexplored territory in the ongoing search for efficient countermeasures against invasive cancer.