Abstract Study question Is there an associated increased risk of preterm birth (PTB) in patients who underwent hysteroscopic septoplasty for subseptate uterus (SSU) with twin gestation through IVF-ET? Summary answer This study suggests an increased risk of PTB in patients, who underwent hysteroscopic septoplasty for SSU, and subsequently conceived twins after IVF-ET treatment. What is known already In general, the incidence of PTB is higher after IVF-ET treatment. In addition, Müllerian uterine anomalies, such as unicornuate uterus, are considered risk factors for PTB, even with singleton pregnancy. Furthermore, historical data suggest that some patients with uterine anomaly may have cervical incompetence. Twin gestation is associated with PTB in approximately 50% of cases. It is assumed that the obstetric outcome of twin gestation after successful septoplasty should be no different from those with normal uterine cavity. The impact of previous septoplasty for SSU on the obstetric outcomes of twin gestation after IVF-ET treatment has not been adequately studied. Study design, size, duration This retrospective cohort study included 377 fertility patients who received treatment between 2006-2019. The study group included 162 patients who underwent hysteroscopic septoplasty for SSU and subsequently conceived dichorionic-diamniotic (DI-DI) twins after IVF-ET treatment (Group 1). The control group included 215 patients who had normal endometrial cavity on hysteroscopy (#109) or Transvaginal 3D ultrasound scan with saline sonohysterogram (#106) and subsequently conceived DI-DI twins after IVF-ET treatment (Group 2). Participants/materials, setting, methods This study was conducted at an infertility clinic affiliated with an academic hospital. Demographic, clinical, and obstetric outcome data were collected for both groups and compared using appropriate statistical methods. PTB, severe PTB, extreme PTB, and peri-viable birth were defined as < 37 weeks, <32 weeks, <28 weeks and between 20 weeks-25 weeks 6 days gestation respectively. Main results and the role of chance There was no significant difference in mean age, BMI and infertility duration or in the incidence of smoking, primary infertility, tubal factors and unexplained infertility between the two groups. There was significantly lower mean baseline FSH levels (p < 0.01), lower incidence of male infertility (p < 0.001) and higher incidence of ovulatory disorders (p < 0.001) and endometriosis (p < 0.05) in Group 1 compared to Group 2. As expected, there was past history of lower parity rate (p < 0.01) and higher miscarriage rate (p < 0.001) in Group 1 compared to Group 2 prior to septum division. There was significantly higher incidence of PTB (72.8% vs 56.3%, p < 0.01), severe PTB (20.3% vs 8.2%, p < 0.01), extreme PTB (8.3% vs 3.9%, p < 0.05), and peri-viable birth (6.3% vs 1.9%, p < 0.05) in Group 1 compared to Group 2. There was no significant difference in cesarean section rate (82.8% vs 87.5%, p = 0.349) between Group1 and Group 2 respectively. There was significantly lower gestational age at birth in weeks (34.0 + 4.0 vs 35.6 + 3.0, p < 0.001), birth weight of the first newborn (2180.5 + 717.9 vs 2436.4 + 606.6, p < 0.001), and the second newborn (2235.2 + 674.2 vs 2398.0 + 537.9, p < 0.05) in grams in Group 1 compared to Group 2. Limitations, reasons for caution Our study has limitations being retrospective in nature. Data on cervical length measurements via ultrasound, if performed, was not available. However, recall bias was prevented by retrieving data from that submitted to the Society of Assisted Reproductive Technology. Additionally, procedural variation was reduced as this was a single-center study. Wider implications of the findings Our study suggests that twin gestation after IVF-ET in patients who underwent successful hysteroscopic septoplasty for SSU may increase the likelihood of PTB. Such patients may require increased surveillance with ultrasound cervical length and signs for preterm labor. Future research should focus on measures to reduce PTB in such patients. Trial registration number Not Applicable