The Children's Sleep Habits Questionnaire (CSHQ) has been used globally as a sound screening tool for sleep problems in children, and it has been increasingly adopted in China. However, its psychometrical properties have not been examined among preschool-aged Chinese children. This study examined the reliability and validity of the CSHQ in urban, preschool-aged Chinese children. Questionnaires were distributed to 3850 kindergarteners and completed by their parents or other guardians. Another 113 CSHQ questionnaires were distributed to examine the test–retest reliability, which was determined by the intraclass correlation coefficient. Internal consistency was evaluated by Cronbach's α. The construct validity was explored by both exploratory and confirmatory factor analysis. Of the 3850 distributed questionnaires, 3324 (86.3%) were returned, and for test–retest questionnaires, 106 of 113 (93.8%) were returned. Cronbach's α ranged from 0.44 (Night-time awakenings) to 0.63 (Daytime sleepiness); test–retest reliability ranged from 0.54 (Night-time awakenings) to 0.76 (Sleep anxiety), with an exceptionally low subscale of 0.38 (Sleep duration). Confirmatory factor analysis showed that the original eight-domain structure was not well suited for preschool-aged Chinese children, and exploratory factor analysis revealed an acceptable eight-factor structure after certain items were deleted and some items were loaded on two factors. Although the results were not optimum, the CSHQ demonstrated basically acceptable reliability, except for Night-time awakenings and Sleep duration subscales, in urban, preschool-aged Chinese children. However, because of factors such as culture and age difference, the CSHQ's structure must be modified to adapt the preschool-aged Chinese children.