OBJECTIVE： To explore the automated intelligent construction and effect of pharmacy intravenous admixture services （PIVAS） in our hospital. METHODS： The main situation of automated intelligent construction and effect of PIVAS in our hospital was introduced， the time and error rate of related links before （Jul.-Sept. 2016） and after （Oct.-Dec. 2016） the automated intelligent construction of PIVAS in our hospital were collected， and its effects were evaluated. RESULTS： Our hospital conducted automated intelligent construction in selecting， labeling， sorting and delivering medicines in PIVAS and established related system and process construction. After automated intelligent construction， time for selecting medicines per medical order was shortened from （6.78±0.87） s to （2.65±0.71） s； time for labeling was shortened from （3.24±0.71） s to （1.41±0.55） s； time for sorting a bag of finished infusion was shortened from （13.37±2.84） s to （5.33±1.72） s； and time for delivering the same amount of infusion shortened from （35.64±4.33） min to （18.12±3.57） min （P＜0.05）. The error rate of selecting medicines was decreased from （2.35±0.59）‰ to  （0.26±0.21）‰； error rate of labeling was decreased from （1.51±0.45）‰ to （0.22±0.10）‰； error rate of misclassifying in a wrong ward was decreased from （3.47±1.02）‰ to （0.17±0.10）‰； and error rate of delivering was decreased from （1.33±0.55）‰ to （0.13±0.11）‰ （P＜0.05）. CONCLUSIONS： The automated intelligent construction in PIVAS of our hospital has improved work efficiency， reduced error rate and potential medication risk and improved management level in PIVAS.