BACKGROUND: Universal prestorage leukoreduction in Canada created the perception that stored red cells (RBCs) are more hemolyzed than their unfiltered predecessors. A pool-split design tested the effects of leukoreduction on hemolysis of stored RBCs. STUDY DESIGN AND METHODS: Two ABO-matched units were pooled, divided, and then processed into leukoreduced (LR) and nonleukoreduced (NLR) units with the Pall LT-WB or RC-PL systems and sampled during standard processing and storage for testing of sterility, counts, hemolysis, and osmotic fragility. RESULTS: Room temperature (RT) filtration of 10 pairs of LT-WB-LR and -NLR units showed significantly different percentage of hemolysis (0.39%) and osmotic fragility (0.643%) at 42 days. Cold-stored and -filtered units (2 days at 4degreesC before processing) were less hemolyzed, but showed a similar proportional decrease of hemolysis in LR units (0.13% vs. 0.25% at 42 days). RBCs from RC-PL systems showed the lowest hemolysis although there was a filtration effect (0.05% vs. 0.12%, 42 days). Osmotic fragility paralleled hemolysis. Segment samples gave inaccurate results. Two-day prefiltration cold storage reduced hemolysis from 0.36 to 0.07 percent (42 days, p < 0.001). RT-LR hemolysis became significantly higher by Day 10 and 4degreesC LR by Day 12. NLR units showed hemolysis by Day 7. LR units filtered cold were less hemolyzed (p < 0.05) than RT-LR but osmotic fragility was unchanged. CONCLUSIONS: LR-RBCs prepared by any of three methods (LT-WB, RT or cold; RC-PL), filtered at 4degreesC, were less hemolyzed during storage than nonfiltered concentrates: 4degreesC leukoreduction is beneficial for RBCs and does not cause hemolysis or enhance fragility.