BACKGROUND: Little is known about red blood cell (RBC) transfusion practices for children hospitalized for a sickle cell vaso-occlusive pain crisis (VOC). We hypothesized that transfusion would be associated with the development of acute chest syndrome (ACS), lower hemoglobin (Hb) concentration, and lack of hydroxyurea therapy. STUDY DESIGN AND METHODS: This is a secondary analysis of all children admitted for a sickle cell pain crisis enrolled in the Magnesium in Crisis (MAGiC) randomized trial; all had HbSS or S-beta0 thalassemia. ACS development and transfusion administration were prospectively collected during the parent trial. All Hb values during the hospitalization were recorded, as was parent report of child receiving hydroxyurea. Relative risks (RRs) of transfusion were compared between groups. RESULTS: Of 204 enrolled children, 40 (19.6%) received a transfusion. Of the 30 children who developed ACS, 22 (73.3%) received transfusions compared to 18 of 174 (10.3%) without ACS: the RR of transfusion in children with ACS was 7.1 (95% confidence interval [CI], 4.4-11.5). Among those without ACS, the lowest Hb was most strongly associated with transfusions: RR was 3.1 (95% CI 2.0 - 4.7) for each 1 g/dL decrease in lowest Hb. In a binary recursive partitioning model for those without ACS, a lowest recorded Hb level of less than 6.3 g/dL was significantly associated with transfusion during admission (p < 0.01). Hydroxyurea use was not associated with transfusions in any analysis. CONCLUSION: ACS increased the RR of transfusion in children hospitalized for VOC sevenfold. In children without ACS, transfusion was associated with lowest Hb concentration, particularly Hb concentration of less than 6.3 g/dL.