Objective: To study the outcomes of children hospitalized for suspected child abuse before and after the implementation of a management protocol in a hospital in Hong Kong. Study period: Two 2-year periods before (1994-1995) and after (2002-2003) the implementation of the protocol in 1998. Methods: This is a retrospective hospital chart review in which the patients' characteristics, the use of laboratory and radiological examination, abuse substantiation and official registrations are compared between the two study periods. Results: There were 109 and 320 patients admitted for evaluation of child abuse for the periods 1994-1995 and 2002-2003, respectively. Children in both periods were similar in sex ratio, proportion of severe forms of child abuse, rates of abuse substantiation and inclusion in the Child Protection Registry. After the implementation of a management protocol, there has been a significant drop in the proportion of children subjected to investigations such as blood counts (86% vs. 16%, p less than 0.001), clotting study (75% vs. 9%, p less than 0.001), and skeletal survey (78% vs. 6%, p less than 0.001). The average length of hospital stay also dropped from 15.3 days to 6.1 days (p less than 0.001). Conclusions: There has been an almost threefold rise in the number of child abuse cases handled at the hospital during the 10-year interval. With the implementation of a management protocol, only a small proportion of children need laboratory investigations or skeletal survey without any drop in abuse substantiation and official registration. The length of hospital stay has also been significantly reduced.