Current concepts of vascular permeability are largely still based on the Starling principle of 1896. Starling's contribution to understanding vascular fluid homeostasis comes from realising that the transport of fluid to and from the interstitial space of peripheral tissues follows the balance between opposing oncotic and hydrostatic pressures. It is presumed that in peripheral tissues fluid is readily filtered from blood to tissues at the arterial/arteriolar side of the circulation and largely reabsorbed at the venular/venous aspect, excess fluid being removed from the tissue by the lymphatic system. This balance is determined particularly by the properties of the vascular barrier. Recent studies have shown that the endothelial glycocalyx, located with a thickness of at least 200 nm on the luminal side of healthy vasculature, plays a vital role in vascular permeability by constituting the vascular barrier together with the endothelial cells themselves. While water and electrolytes can freely pass through the glycocalyx, plasma proteins, especially albumin, interact strongly. Binding and intercalating plasma constituents with the structural elements of the glycocalyx creates the so-called endothelial surface layer. This is the actual interface between flowing blood and the endothelial cell membrane in vivo. The oncotic pressure difference pertinent to fluid homeostasis is not built up between the intravascular and the interstitial tissue spaces, but within a small protein-free zone beneath the glycocalyx surface layer. This explains why perturbation of the glycocalyx leads to a breakdown of both fluid and protein handling in the coronary vascular bed. Preventing damage to the glycocalyx seems to be a promising goal in cardioprotection in many clinical scenarios, including acute ischaemia, hypoxia and inflammation, and chronic vascular disease as in atherosclerosis, diabetes and hypertension.