Perivascular adipose tissue AT is a critical regulator of vascular function, which until recently has been greatly overlooked. Virtually all arteries are surrounded by a significant amount of perivascular adipose tissue, which has long been considered to serve primarily a supportive, mechanical purpose. Recent studies show that both visceral and perivascular fat is a very active endocrine and paracrine source of inflammatory cytokines and adipokines. The latter include beneficial adipocytokines such as adiponectin or so far unidentified adipocyte derived relaxing factor (ADRF) as the presence of perivascular AT may decrease contractile responses to vasoconstrictive agents. However, in pathological states such as obesity, hypertension, diabetes metabolic syndrome and other cardiovascular disorders perivascular tissue becomes dysfunctional and production of protective factors diminishes while detrimental adipocytokines such as leptin, resistin, IL-6, TNF-alpha or IL-17 increases. Moreover the dysfunction of perivascular fat can lead to imbalance between vascular nitric oxide (NO) and superoxide production. Adipokines also regulate immune system as chemokines (such as MIP-1 or RANTES) and induce inflammation with infiltration of T cells and macrophages to the vessel wall. Interestingly central nervous system can affect vascular function through mediation of perivascular adipose tissue dysfunction. In particular sympathetic nervous system endings are present in both visceral and perivascular AT. This powerful relationship between the brain and the vessel can be termed "brain-vessel axis" in which--we propose in the Review--perivascular adipose tissue may take center stage. The role of perivascular fat in the regulation of blood vessels depends on metabolic state, inflammation and clinical risk factors. In health protective and vasorelaxant properties of perivascular AT dominate while in pathology pathogenetic influences including neural stimulation of sympathetic nerve endings or humoral effects of certain hormones and adipocytokines dominates. We propose to term this state "perivascular adipose tissue dysfunction" in similarity to endothelial dysfunction.