The term "blood substitute" is commonly misused when "red cell substitute (RCS) is meant. The ideal RCS should deliver (i) oxygen; (ii) require no compatibility testing; (iii) cause few side effects; (iv) have prolonged storage qualities; (v) persist in the circulation and (vi) be available at reasonable cost. While no drug with all these qualities is on the near horizon, several early generation RCS are approaching submission for licensing, at least for limited indications. Haemoglobin-derived RCS from human, bovine and recombinant sources, as well as perfluorochemicals that dissolve O2 are in different stages of development. While each formulation has its own physical characteristics, biological activities, and adverse reaction profile, all share one characteristic: the physiological consequences of delivering O2 with small molecules is poorly understood and may both account for prob-lems seen in the clinical trials and provide therapeutic opportunities for the cancer patient. Those RCS in phase III trials all have a half-life measured in hours and are unlikely to replace transfusions or drugs that stimulate erythropoiesis for chronic anaemia, but they may play a role (i) in military and civilian trauma as resuscitation solutions, (ii) as a bridge to transfusion when no compatible blood is immediately available, (iii) as an adjunct to the autologous haemodilution management of surgery, or even (iv) in radiation therapy or the management of cancer-related vascular occlusive syndromes.