Objective: To investigate the sedation weaning strategies in critically ill patients with mechanical ventilation in pediatric intensive care unit (PICU) and to explore the effect of different sedative weaning patterns on withdrawal syndrome. Methods: A single-center prospective cohort study was conducted from April 1, 2016 to April 30, 2017. One hundred and twelve patients who required mechanical ventilation and benzodiazepines and (or) opioids for at least 5 consecutive days in PICU of Shanghai Children's Medical Center were enrolled. Twenty patients (17.9%) had an intermittent weaning pattern, defined as a 50% or greater increase in daily benzodiazepine and (or) opioid dose after the start of weaning, and the remaining 92 cases (82.1%) had a steady weaning pattern. The demographic and clinical features, duration and dose of sedative and analgesics, and the incidence of withdrawal syndrome were evaluated. Mann-Whitney U test was used for comparison about clinical features between different weaning pattern groups and children with withdrawal syndrome or not. Logistic regression was used to explore the risk factors of withdrawal syndrome. Results: Among the 112 patients, 46 (41.1%) had withdrawal syndrome. The patients with the intermittent weaning pattern had a high score of pediatric risk of mortality ? (PRISM-?) (10.0 (3.5, 12.0) vs. 6.0 (2.0, 10.0), U=654.50, P=0.043) and were prone to re-intubation (35.0% (7/20) vs. 7.6% (7/92), P=0.003). The patients with withdrawal syndrome had longer duration of sedation (19.5 (16.8, 24.3) vs. 10.0 (7.0, 17.3) days, U=743.50, P<0.01), higher incidence of intermittent weaning pattern (32.6% (15/46) vs. 7.6% (5/66),?(2)=11.58, P=0.001), longer PICU hospitalization (19.0 (15.8, 25.3) vs. 12.0 (8.8, 17.0) days, U=755.00, P<0.01) and higher cost (89 (57,109) vs. 53 (32, 79) thousand yuan, U=804.00, P<0.01). Logistic regression showed that intermittent weaning pattern (odds ratio (OR)=4.85, 95% confidence interval (CI) 1.39-16.91, P=0.013), perioperative period of liver transplantation (OR=6.97, 95%CI 1.25-39.04, P=0.027) and a cumulative dose of midazolam ? 34.7 mg/kg (OR=8.12, 95%CI 3.09-21.37, P<0.01) were risk factors of withdrawal syndrome. Conclusions: Withdrawal syndrome is more likely to occur in children who are intermittently weaned from sedation. Steady weaning strategy may help prevent iatrogenic withdrawal syndrome.