Neovasculature development is a crucial step in the natural history of a cancer. While much emphasis has been placed on proangiogenic growth factors such as VEGF, it is clear that endogenous angiogenesis inhibitors also have critical roles in the regulation of this process. Recent research has identified several cryptic fragments of extracellular matrix/vascular basement membrane proteins that have potent antiangiogenic properties in vivo. It has become apparent that many of these fragments signal via interactions with endothelial integrins, although multiple downstream effector pathways have been implicated and endostatin, the first non-collagenous domain of collagen XVIII, influences an intricate signalling network. The activity of these molecules in animal models suggests that they may have significant clinical activity; however, results of phase I/II trials with endostatin were disappointing. Many possible reasons can be found for the failure of these studies. Weaknesses in trial design, endostatin administration regimen and patient selection are identifiable, and importantly the lack of a clearly defined antiangiogenic mechanism for endostatin hindered assessment of biologically effective dose. Additionally, in vivo immunological and proteolytic function-neutralising mechanisms may have negated endostatin's actions. Lessons learned from these studies will aid the future clinical development of other antiangiogenic extracellular matrix protein fragments.