BACKGROUND: Various procedures can be used to isolate stem and progenitor cells from cord blood. This study evaluated the hydroxyethyl starch sedimentation (HES) with two centrifugation steps, and the top and bottom (T&B) isolation of buffy coat following a single centrifugation, and two filter systems for processing cord blood, one developed by Asahi Kasei Medical (filter A) and the second by Terumo (filter B). METHODS: Each of seven laboratories was randomly assigned the evaluation of either the HES or T&B method and one of the filter methods (n=8 cord blood units, per laboratory, for each method). The leukocyte-containing fraction with the stem/progenitor cells was recovered from the filters by reverse flushing. Utilizing the routine traditional processing and testing procedures of each laboratory, in vitro parameters were determined, with samples obtained after collection, after processing and after freezing/thawing. The results were expressed as the percentage recovery of viable cells in processed vs. collected samples (performance 1; PF1) and in thawed vs. processed samples (performance 2; PF2). The composite results obtained by the seven laboratories were summarized. RESULTS: The median PF1 percentage recovery of total nucleated cells (TNC) was comparable with both traditional methods (HES 79%, T&B 86%) and statistically reduced with both filtration procedures (filter A 58%, filter B 61%). Mononuclear cell (MNC) PF1 recovery was highest statistically with the T&B method (91%) and reduced on using filter A (77%) and filter B (70%) and the HES method (72%). CD34+ cell recovery was judged to be essentially comparable with the four methods, although the range of unit recoveries differed. The percentage recovery of TNC and MNC in PF1 was influenced by the volume of the collected cord blood, especially with use of the filtration procedures. This correlated with TNC content. A greater percentage of red cells and platelets was removed during processing with both filter methods. The time to process cord blood preparations with filter A was significantly shorter than the other methods. Processing with the HES method took the longest time. The recoveries for TNC, MNC and CD34+ cells in PF2 did not appear to be influenced by the specific processing procedure. DISCUSSION: These data indicate that filters that capture stem and progenitor cells may be an appropriate methodology for processing cord blood collected for banking.