OBJECTIVE To investigate the risk factors, clinical prediction and intrapartum management of shoulder dystocia in non-macrosomia. METHODS Totally 7 811 cases of vaginal delivery were retrospectively reviewed from Juanary 2009 to December 2013 in Shengjing Hospital. Shoulder dystocia was found in 11 cases (0.14% , 11/7 811), including 1 case of macrosomia and 10 cases of non-macrosomia (shoulder dystocia group). Each non-macrosomia shoulder dystocia case was matched with 10 cases of normal delivery in the same week, which were selected randomly as the control group. The tendency and risk factors of shoulder dystocia in macrosomia and non-macrosomia were analyzed, and the following data between the two groups were compared, including the height of uterus fundus, abdominal circumference of the pregnant woman, the increasing of body mass index (BMI), fetal biparietal diameter (BPD), fetal femur length (FL), duration of every stage of labor, birth weight of the newborn, head circumference and chest circumference of the newborn, Apgar score. RESULTS (1) There were 213 macrosomias among the 7 811 vaginal deliveries, with the incidence of 2.73% (213/7 811). Only 1 shoulder dystocia was macrosomia (0.46%, 1/213); while the other 10 cases were non-macrosomia ( 0.13%, 10/7 598). (2) From 2009 to 2013, the macrosomia happened by 24 cases (2.32%, 24/1 034), 42 cases (3.61%, 42/1 164), 46 cases (2.60%, 46/1 772), 62 cases (3.01%, 62/2 060), 39 cases (2.19%, 39/1781), respectively. The incidence of macrosomia had no significant difference among these 5 years (P > 0.05). The shoulder dystosia occurrence without macrosia in these 5 years were 1 case ( 0.10% , 1/1 034), 3 cases (0.26%, 3/1 164), 2 cases ( 0.11%, 2/1 172), 2 cases (0.10%, 2/2 060), 2 cases ( 0.11%, 1/1 781), respectively. The incidence of shoulder dystocia without macrosomia had no significant difference among these 5 years (P > 0.05). (3) In the should dystocia group, 5 cases were complicated with premature rupture of membrane (5/10), 4 cases were mother≥ 35 years old (4/10), 3 cases were multipara(3/10), 3 cases had gestational diabetes mellitus(3/10), 3 cases were occiput posterior during the first stage of labor (3/10), 3 cases had prolonged second stage of labor (3/10) and 6 cases had routine lateral incision (6/10). In the control group, 3 cases were complicated with premature rupture of membrane(3/10); 1 case was mother≥35 years old (1/10); 2 cases were multipara(2/10), 3 cases had gestational diabetes mellitus (3/10), 1 case had prolonged second stage (1/10) and 7 cases had routine lateral incision (7/10). (4) There were no significant difference in the height of uterus fundus, BMI, BPD, FL, and duration of the first stage of labor between the shoulder dystocia group and the control group (P > 0.05). Compared with the control group, the increasing of BMI [(6.8±3.1) vs (4.8±1.4) kg/m(2)], the time of the second stage of labor[(86±65) vs (38±28) minutes ] and abdominal circumference[(108±8) vs (101±7) cm] were significantly higher in the shoulder dystosia group (P < 0.05). (5) There were significant difference in the chest circumference of the newborn [(34.0±1.6) vs (32.2±1.9) cm ] and the ratio of chest circumference to head circumference of the newborn [(0.99±0.03) vs (0.97±0.03) ] between the two groups (P < 0.05). The 1-minute Apgar score of the newborn (7.4±2.8) was significantly lower than the control group (10.0±0.0) (P < 0.01). Clavicular fracture occurred in 3 newborns and brachial plexus injury occurred in 4 newborns in the shoulder dystosia group. CONCLUSION It is difficult to predict shoulder dystocia in non-macrosomia. Shoulder dystocia of non-macrosomia could be predicted by measurement of the head circumference, chest circumference, the ratio of chest circumference to head circumference by using prenatal ultrasound. The risk factors may complicated with premature rupture of membrane, abnormal occiput position during the first stage of labor and prolonged second stage of labor.