We evaluated the use of a semi-automated processing technique to salvage red blood cells from pediatric bone marrow donors to minimize the risk of severe anemia following bone marrow harvest and ABO incompatibility in the recipient. Sixty healthy, HLA-matched, pediatric donors of bone marrow hematopoietic cells with a median age 8.0 years (2-19) were studied. Thirteen of the donor-recipient pairs were ABO incompatible. There were 60 recipients with a median age of 8.6 years (2 months to 20.8 years). Bone marrow was harvested under general anesthesia, filtered in the operating room and then transferred to the stem cell laboratory for processing. Samples were obtained for cell count, CD34+ quantification, colony assay, viability, and bacteriologic cultures before and after processing. The cells were processed in a semi-automated closed system (Stericel, Terumo) by density gradient separation with Ficoll-Hypaque and then washed. Two aliquots were obtained: one containing the mononuclear cell layer to be infused to the recipient and the other the washed red cells to be infused to the donor. The median volume harvested was 608 +/- 40.42 ml (278-1409), while the final volume infused was 174 +/- 10.75 ml (30.2-380) P < 0.0001, representing a decrease of 72% of the volume infused. The nucleated cell count harvested was 1.6 x 10(10) +/- 0.1 (0.56-3.2), while the count infused was 6.9 x 10(9) +/- 0.1 (0.12-5.4) P < 0.0001. The median mononuclear cell count (MNC) per kg harvested was 0.67 x 10(8) +/- 0.05 (0.18-2.0) vs an infused cell number of 1.3 x 10(8) MNC/kg +/- 0.1 (0.6-33.6) P < 0. 0001. The CD34+ cells harvested were 2.8 x 10(6)/kg +/- 0.1 (0.25-10.2) vs an infused number of 6.0 x 10(6)/kg +/- 0.5 (0.84-31.0) P < 0.0001. The viability before and after processing was 99%. Red cell salvage performed in a semi-automated closed system is safe and reduces the risk of post-bone marrow harvest anemia in pediatric donors, decreases the volume infused into the donor and enriches the mononuclear and CD34+ cell population, without affecting hematopoietic reconstitution.