Objectives To assess associations between pre-pregnancy BMI and severe maternal morbidity (SMM), perinatal death and neonatal morbidity in twin pregnancies. Design Retrospective cohort study. Setting British Columbia, Canada (2000-2017). Population All twin births at ≥20 weeks’ gestation. Methods We examined associations between pre-pregnancy BMI, categorized as underweight (BMI <18.5m/kg2), normal BMI (18.5m/kg2 BMI <25m/kg2), overweight (25m/kg2 BMI <30m/kg2) and obese (BMI ≥30m/kg2) and adverse maternal and perinatal outcomes, using targeted maximum-likelihood estimation, adjusted for demographics and obstetric history. Outcomes SMM, perinatal death, severe perinatal morbidity. Results Overall, 7770 (368 underweight, 1704 overweight and 1016 obese) women with twin pregnancy were included. The rates of SMM (per 10,000 pregnancies) were: 271.1, 320.4, 270.0 and 225.9 in underweight, normal BMI, overweight and obese women, respectively. Underweight women had higher rates of the composite perinatal adverse outcome (adjusted rate ratio) [aRR] = 1.79, 95%CI = 1.32- 2.43), largely driven by increased rates of severe respiratory distress syndrome, and neonatal death (aRR = 2.81, 95%CI = 1.64-4.83). There was no evidence of elevated risk for perinatal outcomes among overweight and obese women. Conclusions In this population-based study cohort study, we found no evidence of an increased risk of adverse maternal and perinatal outcomes in overweight and obese women with twins; however, infants of underweight women are at higher risk of neonatal death and morbidity. These findings have implications for pregnancy care of women with twin pregnancies, especially those who are underweight. Funding SickKids Foundation (SKF-154852) Keywords: pre-pregnancy obesity, maternal morbidity, perinatal death