Over the last years, the relevance of the matrix metalloproteinase (MMP) family in cancer research has grown considerably. These enzymes were initially associated with the invasive properties of tumour cells, owing to their ability to degrade all major protein components of the extracellular matrix (ECM) and basement membranes. However, further studies have demonstrated the implication of MMPs in early steps of tumour evolution, including stimulation of cell proliferation and modulation of angiogenesis. The establishment of causal relationships between MMP overproduction in tumour or stromal cells and cancer progression has prompted the development of clinical trials with a series of inhibitors designed to block the proteolytic activity of these enzymes. Unfortunately, the results derived from using broad-spectrum MMP inhibitors (MMPIs) for treating patients with advanced cancer have been disappointing in most cases. There are several putative explanations for the lack of success of these MMPIs including the recent finding that some MMPs may play a paradoxical protective role in tumour progression. These observations together with the identification of novel functions for MMPs in early stages of cancer have made necessary a reformulation of MMP inhibition strategies. A better understanding of the functional complexity of this proteolytic system and global approaches to identify the relevant MMPs which must be targeted in each individual cancer patient, will be necessary to clarify whether MMP inhibition may be part of future therapies against cancer.