Abstract: Objective To evaluate the association of pre-pregnancy body mass index (BMI) and adverse pregnancy outcomes. Methods Pregnant women were surveyed in Guangdong Provincial Maternity and Child Care Hospital from January 1, 2009 to January 1, 2012. The content of the survey included the pregnant age, gestational age, gravidity, birth history, past medical history, history of smoking and metabolic diseases (hypertension, diabetes), and family inheritance history. Their weight gains during gestation and pregnancy outcomes were carefully followed up. Results A total of 3 608 women were enrolled in this study. Their average age was（28.25±4.13）years, average pre-pregnancy BMI was（20.34±2.82）, and average growth of gestational weight was （15.21±4.35）kg. The BMI and body weight increased with the increase in gestational weeks (both P<0.01). Among all the women, 669 were underweight (BMI<18.5), 2 476 had normal weight (18.5≤BMI＜24.0), and 463 were overweight or obese (BMI≥24.0) before pregnancy. Incidence rates of adverse outcomes in different groups were 6.73%（45/669）, 33.12%（820/ 2 476）, and 46.87%（217/463）, （P＜0.01 between the three groups）. The proportion differences of outcomes of pregnancy, preeclampsia, GDM, preterm birth, fetal macrosomia, and full-term low birth weight infants were statistically significant (all P < 0.01) for the women with different pre-pregnant weights. Preeclampsia, GDM, premature delivery, and incidence of macrosomia were positively correlated with the pre-pregnancy weights (r=58.37, 85.33, 50.20, 23.62, all P < 0.01). The incidence of full-term low birth weight infants was negatively correlated with pre-pregnancy weights (r= -10.15, P < 0.05). Conclusion Abnormal pre-pregnancy BMI may be closely related to adverse pregnancy outcomes. The pregnancy weight gain should be based on pre-pregnancy BMI with different patterns.