Much of the morbidity and mortality associated with diabetes is primarily attributable to sequelae of microvascular and macrovascular disease. Over the past decade, dramatic progress has been achieved in elucidating the fundamental processes underlying the pathogenesis of these complications. Angiogenic factors in particular now appear to play a pivotal role in the development of microvascular complications as well as the response to macrovascular disease. Hyperglycemia, other growth factors, advanced glycation end products, oxidative stress, and ischemia can increase growth factor expression. In some microvascular tissues, the result is pathologic neovascularization and increased vascular permeability. These responses account for much of the visual loss associated with diabetic retinopathy and may, in addition, serve a significant role in nephropathy and neuropathy. In contrast, recent data suggest that vascular collateralization resulting from ischemia-induced growth factor release in tissues compromised by macrovascular disease may be important in reducing clinical symptoms and tissue damage. This angiogenic response, which may be beneficial in coronary artery and peripheral limb disease, appears to be reduced in patients with diabetes. Thus, two apparently diametrically opposed therapeutic paradigms are arising for the treatment of vascular complications in diabetes. Indeed, growth factor antagonists have been used successfully in diabetes-related animal models to block angiogenic and permeability complications in the retina and kidney. Conversely, growth factor agonists have been successfully used to stimulate collateral vessel formation and reduce ischemic symptoms from macrovascular disease in the coronary arteries and peripheral limbs. Both of these approaches are currently being evaluated in clinical trials for their respective indications. Thus, as these divergent therapeutic modalities begin to enter the clinical arena, this apparent paradox necessitates careful consideration of the potential risks, benefits, and interactions of the opposing regimens. Using vascular endothelial growth factor as a classic example of growth factor involvement, we discuss the current preclinical and clinical data supporting these approaches and the implications arising from the probable coexistence of these two therapeutic modalities.