It has recently become evident that elevation of reticulocytes in the circulation of several species, including humans, leads to the formation of a noncellular transferrin receptor (TFR). In humans, the majority of the released receptor is in truncated form (Shih et al: J Biol Chem 265:19077, 1990). In other species (sheep, rat, chicken) the receptor is associated with a vesicle (exosome) and is full length (Johnstone et al: J Cell Physiol 147:27, 1991). In this report we show that in sheep reticulocytes incubated in vitro, the majority (approximately 75%) of the released receptor is of native size and is exosome associated. A fraction (approximately 25%) is a truncated form of approximately 80 Kd corresponding to the exofacial domain of the TFR. Herein we also address the question of whether the truncated receptor originates by proteolytic cleavage directly from the cell surface or by cleavage from exosomes. Using surface 125I-labeled sheep reticulocytes as the experimental model, we show that during in vitro maturation, 125I-TFR of native size appears in exosomes before the soluble, truncated, exofacial domain of the receptor is detected in the medium. Because cleavage and release of the exofacial domain would likely leave the truncated cytoplasmic and transmembrane domains in the originating membrane (plasma membranes or exosomes), both fractions were probed with antibodies specifically generated against the cytoplasmic domain of the receptor. Only exosomes, not plasma membranes, show the presence of a approximately 17-Kd peptide recognized by the antibody to the cytoplasmic domain of the transferrin receptor. Thus, it is concluded that the truncated, soluble receptor originates from exosomes in sheep. A 17-Kd cytoplasmic domain of the TFR was also detected in exosomes from the reticulocytes of an anemic man, suggesting that the truncated receptor in man may also originate from exosomes. Using in vitro cultures of surface 125I-labeled sheep reticulocytes, it is concluded that exosome formation is the principal route for maturation-associated loss of the TFR. A similar conclusion was made earlier (Johnstone et al: J Cell Physiol 147:27, 1991) for the nucleoside transporter of maturing sheep reticulocytes.