The matrix metalloproteinases (MMPs) are a family of more than 20 distinct enzymes that are frequently overexpressed in human tumors. Functional studies have shown that MMPs play an important role in the proteolytic destruction of extracellular matrix and basement membranes, thereby facilitating tumor invasion and metastasis. In addition, these enzymes may also be important in other steps of tumor evolution including neoplastic cell proliferation and angiogenesis stimulation. On the basis of the relevance of MMPs in tumor progression, a number of different strategies aimed to block the unwanted activity of these enzymes in cancer have been developed. Unfortunately, most clinical trials with the first series of MMP inhibitors have failed to show clear benefit in patients with advanced cancer. Explanations for this lack of success include the failure to recognize the role of these enzymes in early stages of the disease as well as inadequacy of either the employed inhibitors or the proteases to be targeted. The introduction of novel concepts such as tumor degradome, and global approaches to protease analysis, may facilitate the identification of the relevant MMPs that must be targeted in each individual cancer patient. On the other hand, the finding that MMPs are enzymes whose effects on biologically active substrates can have profound consequences on cell behaviour, suggests that selective inhibition of a limited set of MMPs at early stages of tumor evolution might be much more effective than using wide-spectrum inhibitors active against most family members, and administered to patients at late stages of the disease. Further studies directed to elucidate these questions will be necessary to clarify whether any of the multiple strategies of MMP inhibition may be part of future therapeutic approaches to control tumor progression.