A phospholipid vesicle that encapsulates a concentrated hemoglobin (Hb) solution and pyridoxal 5'-phosphate as an allosteric effector [Hb vesicle (HbV) diameter, 250 nm] has been developed to provide an O2 carrying ability to plasma expanders. The O2 release from flowing HbVs was examined using an O2-permeable, fluorinated ethylenepropylene copolymer tube (inner diameter, 28 microm) exposed to a deoxygenated environment. Measurement of O2 release was performed using an apparatus that consisted of an inverted microscope and a scanning-grating spectrophotometer with a photon-count detector, and the rate of O2 release was determined based on the visible absorption spectrum in the Q band of Hb. HbVs and fresh human red blood cells (RBCs) were mixed in various volume ratios at a Hb concentration of 10 g/dl in isotonic saline that contained 5 g/dl albumin, and the suspension was perfused at the centerline flow velocity of 1 mm/s through the narrow tube. The mixtures of acellular Hb solution and RBCs were also tested. Because HbVs were homogeneously dispersed in the albumin solution, increasing the volume of the HbV suspension resulted in a thicker marginal RBC-free layer. Irrespective of the mixing ratio, the rate of O2 release from the HbV/RBC mixtures was similar to that of RBCs alone. On the other hand, the addition of 50 vol% of acellular Hb solution to RBCs significantly enhanced the rate of deoxygenation. This outstanding difference in the rate of O2 release between the HbV suspension and the acellular Hb solution should mainly be due to the difference in the particle size (250 vs. 7 nm) that affects their diffusion for the facilitated O2 transport.