PurposeTo determine the psychometric properties of the Patient-reported Outcomes Measurement Information System Parent Proxy-25 Profile (PROMIS-25) in a sample of Chinese parents of children with cancer. Design and methodsA cross-sectional sample (N = 148) of parents with children aged 5–17 years living with cancer was recruited. Each participant completed sociodemographic and clinical questionnaires and PROMIS-25. The flooring and ceiling effects were calculated. Reliability was determined by the Cronbach's alpha and split-half coefficient. Factor structure was examined by factor analysis. Model fit and graphical plots were assessed to test the assumptions of Rasch model-based item response theory (IRT). Differential item functioning (DIF) was assessed in terms of gender, age, and treatment stage. ResultsPROMIS-25 demonstrated some flooring and ceiling effect, excellent reliability (Cronbach's α > 0.7 for all six domains), and the six-domain factor structure was supported. The IRT assumptions were met in terms of unidimensionality, local independence, monotonicity, and measurement equivalence with acceptable DIF in terms of gender, age, diagnosis, and treatment stage. ConclusionPROMIS-25 is a highly reliable and valid instrument for evaluating children with cancer that assesses important health-related quality of life domains of child cancer. Practice implicationChinese parents of children with cancer and healthcare providers could use PROMIS-25 to assess the symptoms of children.