Collagen, the major constituent of human dermis, represents the main barrier against progression of melanoma cells. Several matrix metalloproteinases (MMPs), i.e. collagenase-1 (MMP-1), gelatinase A (MMP-2) and membrane-type 1-MMP (MMP-14), favor melanoma cell invasion through their capacity of degrading collagen and thus, are considered as main targets. Potent inhibitors, as hydroxamate- derived pseudopeptides were first proposed as pharmacological agents to control melanoma invasiveness. These molecules have major drawbacks linked to i) toxicity and ii) absence of specificity, in keeping with the high Zn chelating property of hydroxamates, that might hinder the contribution of the occupancy of other subsites in enzyme inhibition. To date, research focuses on the design of compounds which display a lower affinity for Zn in enzyme active site. For instance, hydroxamate can be replaced by phosphinic acid or hydrazide which further allows synthesis of both right- and left- hand side inhibitors and therefore occupancy of non-primed enzyme subsites. Novel types of selective MMP inhibitors also include non-competitive and mechanism- based inhibitors. Finally, collagenolysis may be controlled by modulating enzyme-substrate interaction through the identification of substances that bind to MMP exosites. Such compounds could be of value by impeding collagenases to associate to plasma-membrane of invading cancer cells.