Ｏｂｊｅｃｔive　To investigated the effects of preoperative application of ulinastatin(UTI) on serum inflammatory factors and postoperative delirium in elderly patients receiving elective lower limb arthroplasty.　Methods　A prospective, double-blind, randomized controlled study was conducted. A total of 80 elderly patients(65－85 y old) with ASA grades of Ⅰ－Ⅲ were enrolled in elective lower limb arthroplasty. They were randomly and evenly assigned into ulinastatin treated- (UTI) and non-treated- (control) groups (n=40). Before surgery, 300 thousand units of UTI was added into 100 ml normal saline and was intravenously infused in 30 min in UTI group, while 100 ml normal saline was intravenously administered in control group in 30 min. All patients received tracheal intubation under general anesthesia during the operation and were subjected to iliac fascia space block guided by ultrasound. We measured minimum mental state examination(MMSE) scores 1 day before operation and assessed the existence of delirium with confusion assessment method(CAM) at postoperative 24 h(T2), 48 h(T3), and 72 h(T4). We collected peri- and post-operative information from the patients, including demographic information, intraoperative parameters (such as, duration of anesthesia and operation, volume of intraoperative bleeding and blood transfusion, dosage of anesthetics, etc.) and postoperative indices(VAS, postoperative recovery and adverse effects, etc.). We also collected 3 ml peripheral blood from patients before operation(T0) and at post-operative time points (T1, T2 and T3) to detect the serum proinflammatory cytokines, such as, TNF-α, IL-1β, IL-6, and S100β.　Results　The CAM scores in UTI group[(12.8±2.1), (13.1±2.9)] were significantly lower than those of control group[(14.0±2.5), (14.7±3.1)] at T2 and T3(P<0.05), respectively. The prevalence of postoperative delirium was not significantly different at T2 through T4 between two groups(P＞0.05). S100β reached peak at T1 in both UTI and control groups. Compared with the control group, S100β in serum was significantly lower at T1 in UTI group[(304±142) ng/L vs (396±208) ng/L](P<0.05). In two groups of patients, IL-1β and IL-6 in serum at T0-T2 showed upward trends and reached peaks at T2. Compared with the control group, IL-1β and IL-6 in UTI group decreased significantly at T1 [IL-1β: (70.4±7.3) ng/L vs (81.6±8.5) ng/L, IL-6: (55±8) ng/L vs (62±8) ng/L] and T2 [IL-1β: (99.6±7.7) ng/L vs (105.7±8.0) ng/L, IL-6: (88±8) ng/L vs (99±11) ng/L](P<0.05).　Conclusions　Preoperative treatment with UTI can improve post-operative CAM scores and cognitive function in elderly patients receiving elective lower limb joint replacement. The effects of UTI may be related to its inhibition of inflammatory cytokines.