Vertebrate red blood cells (RBCs) seem to serve tissue oxygen delivery in two distinct ways. Firstly, RBCs enable the adequate transport of O(2) between respiratory surfaces and metabolizing tissues by means of their high intracellular concentration of hemoglobin (Hb), appropriate allosteric interactions between Hb ligand-binding sites, and an adjustable intracellular chemical environment that allows fine-tuning of Hb O(2) affinity. Secondly, RBCs may sense tissue O(2) requirements via their degree of deoxygenation when they travel through the microcirculation and release vasodilatory compounds that enhance blood flow in hypoxic tissues. This latter function could be important in matching tissue O(2) delivery with local O(2) demand. Three main mechanisms by which RBCs can regulate their own distribution in the microcirculation have been proposed. These are: (1) deoxygenation-dependent release of ATP from RBCs, which stimulates production of nitric oxide (NO) and other vasodilators in the endothelium; (2) release of vasoactive NO from S-nitroso-Hb upon deoxygenation; and (3) reduction of naturally occurring nitrite to vasoactive NO by deoxygenated Hb. This Commentary inspects all three hypotheses with regard to their mechanisms, experimental evidence in their support and details that remain unresolved. The prime focus is on human/mammalian models, where most evidence for a role of erythrocyte ATP and NO release in blood flow regulation have accumulated. Information from other vertebrate groups is integrated in the analysis and used to discuss the evolutionary origin and general relevance of each hypothesis.