Objective <p>To investigate risk factors for prolonged intensive care unit (ICU) stay after surgery for congenital oesophageal atresia (OA), with or without tracheo-oesophageal fistula (TOF).</p> Methods <p>This retrospective study enrolled infants who underwent surgical repair for congenital OA. Patients were divided into two groups (prolonged and control group) based on the length of their ICU stay. A prolonged ICU stay was defined as exceeding the 75th percentile of the cohort. Multivariable logistic regression analysis was employed to identify the independent risk factors for a prolonged ICU stay. A receiver operating characteristic (ROC) curve was then used to evaluate the predictive value of these risk factors.</p> Results <p>Between June 2017 and September 2024, 177 infants underwent OA/TOF repair. Of these, 173 met the inclusion criteria after exclusions were applied. The median ICU stay was 13 days (IQR 9–24 days), with 44 patients classified as having a prolonged ICU stay. Univariate analysis identified 15 potential predictors of ICU stay, while multivariate logistic regression analysis revealed 3 independent risk factors for prolonged ICU stay: gestational age (OR: 0.705, 95% CI: 0.594–0.837, <i>p</i> &lt; 0.001), the presence of other congenital anomalies (OR: 6.087, 95% CI: 2.564–14.451, <i>p</i> &lt; 0.001), and early postoperative severe hypoalbuminaemia (OR: 5.183, 95% CI: 1.494–17.979, <i>p</i> = 0.01). ROC curve analysis showed that combining these factors provided a good prediction of prolonged ICU stay after surgery, with an AUC of 0.835 (95% CI: 0.758–0.913).</p> Conclusions <p>This study identified the independent risk factors for a prolonged ICU stay following OA/TOF repair. These results may enable clinicians to implement earlier interventions to reduce risks and prevent delayed recovery.</p>