The contribution of the lymphatics to the absorption and systemic availability of recombinant human epoetin alfa (rHuEPO) following s.c. injection was examined using a cannulated sheep model. Parallel studies were conducted in sheep where a single bolus dose was administered either by i.v. (10, 100, or 1000 IU/kg) or s.c. (400 IU/kg) injection. The first s.c. group served as a control for the calculation of absolute bioavailability. In the second group, the efferent popliteal lymphatic duct was cannulated and peripheral lymph draining the injection site was continuously collected. In the third group, the thoracic duct was cannulated to allow collection of central lymph just prior to entry into the systemic circulation. Blood was periodically sampled from all animals, and concentrations in serum and lymph were determined by enzyme-linked immunosorbent assay. The cumulative amount of rHuEPO recovered in peripheral and central lymph was 83.9 Ϯ 6.6% and 75.3 Ϯ 3.9% of the administered dose, respectively, indicating almost complete absorption from the s.c. injection site and minimal clearance during transit through the lymphatic system. After i.v. administration, the systemic clearance of rHuEPO decreased with increasing dose, reflecting capacity-limited elimination kinetics. A pharmacokinetic model was developed to simultaneously fit experimental data for all treatment groups and estimate bioavailability. The direct measurement of Ͼ75% of the dose in peripheral and central lymph independently verifies the calculated bioavailability of 87% and demonstrates the major role of the lymphatic route in the overall s.c. bioavailability of rHuEPO after s.c. administration with this animal model. Erythropoietin (EPO) is a renally synthesized 30.4-kDa glycoprotein involved in the regulation of red blood cell proliferation and differentiation. Recombinant human epoetin alfa (rHuEPO) has been extensively utilized in the treatment of anemia associated with chronic renal failure, AIDS, and chemotherapy.Pharmacokinetic studies of rHuEPO have demonstrated dose-dependent disposition in humans (Flaharty et al., 1990;Veng-Pedersen et al., 1995), monkeys (Ramakrishnan et al., 2003), sheep (Veng-Pedersen et al., 1999), and rats (Kato et al., 1997) after i.v. administration. In these studies, nonlinear kinetics were characterized by decreased clearance with increasing dose. Although the exact elimination pathway has not been fully elucidated, studies comparing clearance (CL), mean residence time (MRT), and terminal half-life (t 1/2 ) preand postablation of the bone marrow in sheep, have provided significant evidence that elimination occurs predominantly in the bone marrow (Chapel et al., 2001). The capacity-limited CL is likely to be a receptor-mediated endocytosis by erythroid progenitor cells and has been described previously using a Michaelis-Menten-type elimination (Kato et al., 1997;Veng-Pedersen et al., 1999;Ramakrishnan et al., 2003). Although EPO receptors appear to have an important role in the distribution of EPO outside...