The primary mechanism of erythropoietin's (EPO) in vivo elimination and the tissue, or tissues, responsible are unknown. Previous studies indicating that EPO pharmacokinetic (PK) behaviour is nonlinear suggest that EPO elimination takes place by a saturable mechanism. A versatile PK system analysis, the Disposition Decomposition Analysis (DDA), capable of quantification of the Michaelis -Menten parameters, V m and k m was used to analyze and compare EPO's PK behaviour in newborn sheep and preterm infants. Lambs and infants both demonstrated nonlinear PK behaviour appropriately analyzed with DDA. Compared to preterm infants, lambs had significantly greater (pB 0.05) elimination capacity as determined by the V m (27899525 versus 1767 9 250 mU/mL per h (mean9S.E.), respectively), and larger extrapolated linear clearances (116 919.1 versus 21.3 91.75 mL/kg per h, respectively) (p B 0.01). Lambs also demonstrated significantly larger (pB0.01) degrees of nonlinearity as judged by smaller mean k m values (21429258 versus 679691.007 mU/mL, respectively). Of note, although the DDA does not distinguish what the mechanism of EPO elimination is, enzymatic degradation and receptor-mediated cellular internalization are two possibilities. The in vivo DDA-derived k m values were similar to reported in vitro binding affinity k d data for erythroid progenitors and cell lines having EPO-R's, i.e. 240 -2400 mU/mL. The present study's demonstration that EPO's nonlinear PK behaviour in both sheep and humans can be analyzed by the DDA methodology indicates that the sheep model may be used in invasive studies needed to further characterize the mechanism of EPO elimination.