BackgroundQuality of parent–child interaction in early childhood functions as a critical indicator of nurturing care and is strongly associated with short‐term and long‐term development (health, cognition, language, social emotion, well‐being, etc.). NCAST PCI Teaching Scale (PCI‐TS), a video‐based assessment regarded as a gold standard to measure PCI, has been widely used worldwide. However, its psychometric soundness among the urban Chinese population is unclear. This study assesses the PCI‐TS's reliability and validity and explores predictive factors among urban Chinese parent–child dyads.MethodsPCI‐TS was adopted to code mother–child interaction among urban Chinese dyads recruited during the children's regular health checks in local maternal and child health centres. Reliability was evaluated by internal consistency (Cronbach'α coefficient) and test–retest reliability (Pearson correlation) with an average interval of 18 days. Score distribution of each subscale and total scale were compared with NCAST Database and Canadian community sample by single sample t‐test. Criteria‐related validity was conducted by Infant‐Toddler Home Observation Measurement of the Environment (Pearson correlation). Predictive factors was performed by multiple linear regression.ResultsFour hundred and twenty‐nine eligible mother–child dyads were included for data analysis among the 466 recruited samples. Four qualified local paediatricians accomplished video coding with an average agreement of 86%. The PCI‐TS has strong reliability among the Chinese population with the Cronbach'α coefficients of the Caregiver‐Infant total score, Caregiver total and Infant total scores of 0.81, 0.81 and 0.74, respectively; an acceptable test–retest reliability (r = 0.73, p < 0.01); and moderate correlation with IT‐HOME, ranging from 0.53 to 0.62. Child age, birth weight, maternal education, full‐time housewife, living with grandparent(s) and living space were predictive factors on PCI‐TS in the Chinese population.ConclusionPCI‐TS showed good psychometric properties for measuring mother–child interactions among urban Chinese dyads, offering clinicians and researchers a practical tool to evaluate PCI objectively. Child age, maternal education and living space were beneficial factors, while full‐time mothers and living with grandparent(s) were risk factors.