The existence of a local cardiovascular renin-angiotensin system (RAS) is often invoked to explain the long-term beneficial effects of RAS inhibitors in heart failure and hypertension. The implicit assumption is that all components of the RAS are synthesized in situ, so that local angiotensin II formation may occur independently of the circulating RAS. Evidence for this assumption however is lacking. The angiotensin release from isolated perfused rat hearts or hindlimbs depends on the presence of renal renin. When calculating the in vivo angiotensin production at tissue sites in humans and pigs, taking into account the extensive regional angiotensin clearance by infusing radiolabeled angiotensin I or II, it was found that angiotensin production correlated closely with plasma renin activity. Moreover, in pigs the cardiac tissue levels of renin and angiotensin were directly correlated with their respective plasma levels, and both in tissue and plasma the levels were undetectably low after nephrectomy. Similarly, rat vascular renin and angiotensin decrease to low or undetectable levels within 48 h after nephrectomy. Aortic renin has a longer half life than plasma renin, suggesting that renin may be bound by the vessel wall. In support of this assumption, both renin receptors and renin-binding proteins have been described. Like ACE, renin was enriched in a purified membrane fraction prepared from cardiac tissue. Binding of renin to cardiac vascular membranes may therefore be part of a mechanism by which renin is taken up from plasma. It appears that the concept of a local RAS needs to be reassessed. Local angiotensin formation in heart and vessel wall does occur, but depends, at least under normal circumstances, on the uptake of renal renin from the circulation. Tissues may regulate their local angiotensin concentrations by varying the number of renin receptors and/or renin-binding proteins, the ACE level, the amount of metabolizing enzymes and the angiotensin receptor density.