Objectives: Despite the fact that many published guidelines for optimal usage of blood products are available, RBC transfusions are frequently guided by personal preference and tradition. The purpose of the study was to examine the appropriateness of RBC transfusions for surgical patients admitted to the SICU and how it might affect the clinical outcomes. Methods: We retrospectively reviewed all RBC transfusions that occurred in a general SICU between July 1, 2015 and June 30, 2016. Each RBC transfusion was critically ap-praised using four criteria to determine whether or not the transfusion was indicated. De-mographic data and clinical outcome measures of transfused and non-transfused patients as well as patients with indicated or non-indicated RBC transfusions were compared. Results: There were 617 patients admitted to the SICU and 245 (39.7%) of them required RBC transfusions. Transfused patients were more likely to have had operations prior to ICU ad-mission (87.3% vs. 75.8%) and have higher APACHE II scores (22.1 ± 8.5 vs. 15.5 ± 7.4). They also had longer ICU stays (8.0 ± 9.1 vs. 3.0 ± 3.7 days) and higher ICU mortality rates (14.3% vs. 6.2%). Among 651 transfusions, 38.4% of them fulfilled at least one crite-rion for transfusion and 61.6% were not considered appropriate. Patients who had non-indicated postoperative RBC transfusion were more likely to have had intraoperative blood loss ≥ 500 mL (p < 0.05). Conclusions: Our efforts to reduce unnecessary RBC transfusion in the SICU appear inadequate. Educating medical personnel is important, but additional measures should also be taken in order to improve compliance.