Metalloproteinases such as the matrix metalloproteinases (MMPs) and disintegrin-metalloproteinases with thrombospondin motifs (ADAMTSs) have been implicated in the pathological destruction of joint tissues in rheumatoid arthritis and osteoarthritis. These enzymes degrade extracellular matrix macromolecules and modulate factors governing cell behavior. They may also be involved in tissue repair, but become a part of the destructive disease process due to overexpression. Studies investigating the roles of metalloproteinases have thrown light on the failure of the early clinical trials of MMP inhibitors as therapeutic agents in arthritic diseases. It is now clear that a more accurate knowledge of the enzymes in the different cells and their precise roles in the disease process is required for these approaches to be successful. The next generations of metalloproteinase inhibitors should have added specificity, gained from an understanding not only of the catalytic domain structures but the role of extracatalytic motifs in substrate binding, or by the generation of engineered tissue inhibitors of metalloproteinases. Inhibition of the enzymes by modulating gene expression or preventing protein activation could also be considered. Work on the development of effective biomarkers is also essential before an effective evaluation of the new generations of specific inhibitors can be made.