OBJECTIVES: This study aims to compare clinical efficacy and costs between insulin analogue and human insulin in treatment of patients with diabetes in a tertiary hospital in Beijing. METHODS: Data were extracted from inpatient electronic patient records in HIS system in a tertiary hospital in Beijing during the period between Aug 2011 and Nov 2013. Inclusion criteria: main admitting diagnoses were type I diabetes or type II diabetes; insulin analogue or human insulin were used during hospital stay. Exclusion criteria: combination use of insulin analogue and human insulin. Fasting blood-glucose (FPG) before breakfast was used as efficacy index while total inpatient expenditure was used as cost index. Unfortunately, HbA1c level was unavailable to this study. Student-t test and Chi-square test were conducted using SPSS 20.0. RESULTS: Insulin analogue (n=50) and human insulin (n=49) cohort had comparable male (68% vs 49%, P=0.055), age (50 vs 57, P=0.008), and baseline clinical characteristics (length of stay 10 vs 11, P=0.133; baseline FPG 8.73mmol/L vs 8.44mmol/L, P=0.630). At the point of discharge, FPG before breakfast was dropped to 6.39mmol/L vs 7.13mmol/L (P=0.025) in insulin analogue and human insulin cohort, respectively. The total medical costs were CNY 11,305 vs CNY 10,693 (P=0.577), and total drug costs were CNY 5,198 vs CNY 4,199 (P=0.186), in insulin analogue and human insulin cohort, respectively. CONCLUSIONS: Insulin analogue treated patients experienced significantly greater reductions in FPG before breakfast compared to human insulin treated patients, while total inpatient expenditure and drug costs showed no significant difference between the two cohorts. Should data permits, HbA1c data should be included in further analysis in the future.