Inhibition of matrix metalloproteinases (MMPs), a family of proteolytic enzymes linked to many aspects of cancer progression, has been explored as a therapeutic goal for almost two decades. Thus far, all tested MMP inhibitors (MMPIs) have failed to reach primary end points in Phase III clinical trials, although secondary analyses suggest benefits in particular patient groups. The clinical development of these agents has been hampered by problems related to determination of effective dosages and side effects that necessitate dose lowering or drug holidays. Imaging technologies offer hope as a means to measure enzyme activity and hence effective enzyme inhibition in vivo. Meanwhile, recent results from genetic studies of both mice and man have given some clues to possible causes of musculoskeletal side effects. Future progress in the therapeutic use of MMPIs is dependent on the ability to selectively target cancer-associated MMPs at the correct stage in tumour progression and the development of surrogate markers of in vivo efficacy.