Objective: This study aimed to survey the prevalence of delirium in the pediatric intensive care unit (PICU) and explore the associated risk factors. Design: A retrospective case-control study. Setting: Two PICUs within a tertiary-A general hospital. Patients: Patients aged from 1 month to 7 years who stayed in either PICU for at least 1 day were included. Methods: A total of 639 patients admitted to PICU of a tertiary-A general hospital from December 2018 to August 2019 were enrolled. Demographic, clinical, laboratory data and length of stay in the PICU were collected. The patients were screened twice a day with the Chinese version of Cornell Assessment of Pediatric Delirium (CAPD), and were divided into the delirium group and the non-delirium group. A risk factor analysis was conducted, with ICU pediatric delirium as primary outcome, by performing a multivariable logistic regression analysis. Results: Of the 639 patients, the prevalence of ICU pediatric delirium was 31.30%. Of the 200 children with delirium across 3703 study days, 36% children were hyperactive, 41% were hypoactive, and 23% displayed the mixed type of delirium. Univariate analysis and multivariate logistic regression analysis showed that age, PRISM IV score (OR, 2.20; 95% CI, 1.42-3.41), hypoxia (OR, 2.69; 95% CI, 1.53-4.71), metabolic dis-function (OR, 3.73; 95% CI, 2.08-6.71), duration of infection (OR, 1.22; 95% CI, 1.10-1.36), and mechanical ventilation (OR, 3.78; 95% CI, 2.25-6.35) were statistically correlated with ICU pediatric delirium. The ROC curve analysis shows the combination CRP with duration of infection has good predictive performance. Conclusions: Age, PRISM IV score, ICU retention time, metabolic dis-function, duration of infection, hypoxia, CRP and mechanical ventilation were the independent risk factors for ICU pediatric delirium. We suggest that active preventive measures should be taken to reduce the occurrence of ICU pediatric delirium.