Background Vitamin A can have a positive impact on growth and development of children, but vitamin A deficiency (VAD) was found to be a public health problem in Zhejiang Province, China in 1998. There have been no studies on this topic in Zhejiang Province recently. This study was designed to evaluate the serum retinol levels of children aged 0~4 years in Zhejiang Province, southeast China. This epidemiological data will help design supplementation strategies for vitamin A in high-risk groups and improve their vitamin A status. Methods Children were randomly recruited for this study using a stratified sampling method. A blood sample was collected from each child. Assessment included C-reactive protein (CRP), serum retinol was measured with HPLC and a questionnaire completed providing for family information and nutritional status. Logistic regression analysis was used to evaluate the risk factors for VAD in children. Results A group of 357 subjects aged 1 day to 4 years were recruited. The mean plasma retinol concentration was 1.653 (sd 0.47) μmol/L. There were 3.08% (11/357) of children affected with VAD, and 7.28% (26/357) of children had low vitamin A status, but none of the children showed any clinical symptoms of VAD. There was no significant difference in the levels of plasma retinol and the incidence rate of VAD between male and female children. Multivariate logistic regression analysis showed that living in urban region, having parents with good education and taking vitamin A capsule regularly prevented children from VAD, whereas being young (less than 2 years old) was a risk factor. Conclusion Low vitamin A status remains a nutritional problem in Zhejiang Province. The high-risk group in this study were young, dwelled in rural regions, had parents with poor education and did not take a regular vitamin A containing supplement.