Background: Restrictive blood transfusion is recommended by major guidelines for perioperative management, but requires objective assessment at 7-10 g/dl haemoglobin (Hb). A scoring system that considers the physiological needs of the heart may simply the practice and reduce transfusion.Methods: Patients (14-65 years of age) undergoing non-cardiac surgery were randomised at a 1:1 ratio to a control group versus a Perioperative Transfusion Trigger Score (POTTS) group. POTTS (maximum of 10) was calculated as 6 plus the following: adrenaline infusion rate (0 for no infusion, 1 for ≤0.05 μgÁkg À1 Ámin À1 , and 2 for higher rate), FiO 2 to keep SpO 2 at ≥95% (0 for ≤35%, 1 for 36%-50%, and 2 for higher), core temperature (0 for <38 C, 1 for 38-40 C, and 2 for higher), and angina history (0 for no, 1 for exertional, and 2 for resting). Transfusion is indicated when actual Hb is lower than the calculated POTTS in individual patients. Transfusion in the control group was based on the 2012 American Association for Blood Banks (AABB) guideline. The primary outcome was the proportion of the patients requiring transfusion of allogeneic red blood cells (RBCs) during the perioperative period (until discharge from hospital), as assessed in the intention-to-treat (ITT) population (all randomised subjects).Result: A total of 864 patients (mean age 44.4 years, 244 men and 620 women) were enrolled from December 2017 to January 2021 (433 in the control and 431 in the POTTS group). Baseline Hb was 9.2 ± 1.8 and 9.2 ± 1.7 g/dl in the control and POTTS This study was registered at http://www.chictr.org.cn (#ChiCTR-INR-17014085).