Extremely preterm infants lack an appropriate erythropoietic response to anemia, failing to increase erythropoietin (Epo) production appropriately despite a need for improved tissue oxygenation. We hypothesized that the anemia of prematurity might involve a delay in the premature kidney to produce Epo, or possibly an inability of the preterm kidney to respond to anemic hypoxia. To test our hypothesis, we first evaluated mid-trimester human kidneys for the presence of Epo mRNA and Epo protein. Fetal liver and kidney samples were obtained at 11 through 22 weeks gestation. Hep3B cells (known to produce Epo) were used as positive controls. Epo mRNA was found in all kidney and liver samples evaluated between 11 and 22 weeks. Epo mRNA isolated from fetal kidneys appeared relatively equivalent (using semi-quantitative RT-PCR) throughout the gestations evaluated. Immunocytochemistry was performed on 4 micron slices of paraffin-embedded fetal kidney using a monoclonal rabbit anti-human Epo antibody, and detected with an avidin-biotin secondary antibody system. Positive staining could be identified within a sub-population of tubular cells located in the renal cortex. We conclude that Epo message and protein are present in the human kidney during the second trimester. We next determined whether the anemia of prematurity involved a decreased responsiveness of the premature kidney to produce Epo under hypoxic conditions. Fetal tissue was collected and processed within one hour of termination. Minced tissue samples were incubated in serum-free culture media for 24 hours at 21% or 1% oxygen. The supernatant was collected for Epo protein measurement by ELISA. Total protein was measured using a Bradford Assay to compare Epo protein accumulation with total protein accumulation. Although total protein accumulation increased following hypoxic stimulation, Epo accumulation did not increase in fetal kidney samples. Epo accumulation increased 10 to 20-fold in Hep3B cells, and the percent of total protein comprised by Epo increased from <1% to an average of 10%. Under these in vitro conditions, it appears that the human fetal kidney does not increase Epo protein in response to hypoxic stimulation. Whether renal Epo production (in the mid-trimester fetus or in the extremely preterm infant) is biologically significant remains to be determined. We speculate that the anemia of prematurity might result from delayed changes in gene regulation that are governed by gestational age.