The oxygen delivery capacity of fresh whole blood was compared with that of 21-day-old blood in an isolated, perfused canine hindlimb preparation. At the outset, 21-day-old blood was found to have depressed levels of 2,3 DPG and higher p50's relative to those of freshly procured blood, implying an increased oxygen affinity of the former product. Despite this, oxygen consumption from 21-day-old blood was not demonstrated to be significantly different than that from fresh blood. It is concluded that significant improvement in oxygen delivery capacity alone should not be construed as justification for the choice of fresh over older blood clinical replacement therapy.