Epoetin (recombinant human erythropoietin, EPO) is of proven benefit in the treatment of renal anaemia, and preliminary reports suggest that it may have a role in the management of other anaemic conditions. Pharmacokinetic and therapeutic studies have examined the use of epoetin administered intravenously, intraperitoneally and subcutaneously, and there is accumulating evidence that the last route has several advantages. After intravenous administration, epoetin is distributed in a volume comparable to the plasma volume, and plasma concentrations decay monoexponentially with a half-life of between 4 and 12 hours. Administration of epoetin in peritoneal dialysis fluid results in detectable concentrations in the bloodstream after 1 to 2 hours, and peak concentrations of the order of 2 to 4% of those obtained with the same intravenous dose are found after approximately 12 hours. The bioavailability of epoetin administered intraperitoneally in dialysis fluid is about 3 to 8%, but this may be increased by injecting the drug into a dry peritoneal cavity. Subcutaneous administration results in peak concentrations at about 18 hours which are 5 to 10% of those found after the same intravenous dose. The bioavailability of subcutaneous epoetin is about 20 to 30% and detectable serum concentrations are still present 4 days after administration, in contrast to intravenous administration after which concentrations have returned to baseline within 2 to 3 days. Remarkably little is known about the metabolic fate of either erythropoietin or epoetin. In addition, there is much controversy surrounding the relative roles of the kidney and liver in the catabolism of epoetin. About 3 to 10% of epoetin is excreted unchanged in the urine. In common with other glycoproteins, the carbohydrate residues which constitute 40% of its molecular size are essential for maintaining the stability of epoetin in circulation. Desialated epoetin, although biologically active in vitro, is cleared very rapidly from plasma with resultant loss of activity. Further work is required, however, in identifying the pathways of metabolism and elimination of this glycoprotein hormone.