OBJECTIVE： To investigate the relationship of drug utilization index （DUI） of antibiotics with therapeutic efficacy of patients with acute exacorbation of chronic obstructive pulmonary disease （AECOPD）. METHODS： By the method of drug utilization evaluation， inpatients with AECOPD in our hospital during 2013-2015 were selected as the research object. Diagnosis and treatment information prescribing information of patients were collected， and inpatients only receiving one kinds of antibiotics during hospitalization stay were selected to evaluate the relationship of therapeutic efficacy with rehospitalization indexes. RESULTS： A total of 2 155 AECOPD patients were enrolled in the study. Among antibiotics with DDDs＞500， antibiotics with DUI close to 1.0 was cefotiam hydrochloride for injection， and that with DUI far higher than 1.0 was Cefodizime sodium for injection， while that with DUI much less than 1.0 was Piperacillin sodium and sulbactam sodium for injection. There was statistical significance in therapeutic efficacy between Cefotian hydrochloride for injection and Piperacillin sodium and sulbaactam for injection （P＜0.05）； rehospitalization had no statistical significance （P＞0.05）. There was no statistical significance in therapentic efficacy or rehospitalization between cefotiam hydrochloride for injection and cefodizimes sodium for injection （P＞0.05）. CONCLUSIONS： DUI is associated with therapeutic efficacy， but the rationality of antibiotics can not be simple judged according to the distance between DUI and 1.0. A variety of confounding and bias factors should be integrated to avoid misreading and misjudgment.