OBJECTIVE： To evaluate the effects of interventions on the use of human serum albumin in respiratory department of our hospital， and to promote rational clinical drug use. METHODS： According to package inserts， related guideline and literatures of human serum albumin， the evaluation standard of human serum albumin use in respiratory department of our hospital was established with evidence-based medicine method. The comprehensive intervention of human serum albumin use was carried out by means of medical order review， prescription comment， permission setting of information system and performance evaluation. The prescriptions of patients receiving human serum albumin were collected during Jan.-Jul. 2015 （pre-intervention group， 226 cases） and Jan.-Jul. 2016 （post-intervention group， 148 cases）. Those prescriptions were analyzed statistically in respects of total amount of human serum albumin， per capita amount， average duration， indications， contraindications and rationality， etc. The utilization of human serum albumin was compared between 2 groups before and after intervention. RESULTS： The total amount of human serum albumin， per capita amount and average duration decreased from 18 350 g，（81.2±71.3）g，（7.8±7.2）d before intervention to 7 584 g，（51.2±48.3）g，（4.7±2.1）d after intervention， respectively. The proportion of not in line with the indication stated in Guidelines for Use of Human Serum Albumin， Non-protein Colloidal Solution and Srystalline Solution， the proportion of being contrary to contraindication decreased from 80.1％， 50.0％ before intervention to 39.9％， 16.9％ after intervention， respectively. The number of patients with serum albumin concentration ≥35 g/L， ranged ≥30-＜35 g/L or ranged ≥20～＜25 g/L before medication in pre-intervention group was more than post-intervention group； the number of patients with serum albumin concentration＜20 g/L or ranged 20-25g/L in post-intervention group was more than pre-intervention group. The proportion of patient who took human serum albumin rationally increased from 16.8％ before intervention to 52.7％ after intervention （P＜0.05）， there were statistical significances in above-mentioned indexes between before and after intervention. CONCLUSIONS： The comprehensive intervention can effectively reduce the indication， medication course and proportion of unreasonable use of human serum albumin， and improves rational use of human serum albumin.