Background Red blood cells (RBCs) cryopreserved in glycerol must be deglycerolized prior to transfusion. The adequacy of glycerol removal is commonly assessed by measurement of the refractive index (RI) of the supernatant fluid. However, the presence of free hemoglobin in the supernatant falsely increases the RI and may lead to discard of units that have an acceptable residual glycerol concentration. Methods We performed an analysis of the diagnostic accuracy of 3 methods for residual glycerol measurement – refractometry, osmometry, and a glycerol assay kit. Residual glycerol measurement using these methods was performed on 12 deglycerolized, citrate-phosphate-dextrose (CPD)/saline-adenine-glucose-mannitol (SAGM) leukoreduced RBCs. A calculation that estimates the glycerol concentration based on the refractive index and supernatant hemoglobin concentration was developed and ensures that units with an elevated RI due to the presence of hemoglobin are not discarded if their residual glycerol concentration was < 1.0% (w/v). Results Osmometry was an accurate method for estimating residual glycerol concentration. Refractometry overestimated the residual glycerol concentration due to the interference from hemoglobin. However, when supernatant hemoglobin values were measured and used in the calculation for glycerol concentration, refractometry accurately estimated the residual glycerol concentration. Conclusion The residual glycerol concentration of cryopreserved, deglycerolized CPD/SAGM RBCs can be accurately estimated using the supernatant refractive index and an equation that accounts for the supernatant hemoglobin concentration.