OBJECTIVE: This study examined the validity of primary health care providers' (PHCPs) assessment of suspicion that an injury was caused by child abuse and their decision to report suspected child abuse to child protective services (CPS). METHODS: By using a subsample of injuries drawn from the 15,003 childhood injuries evaluated in the Child Abuse Recognition and Evaluation Study, PHCPs completed telephone interviews concerning a stratified sample (no suspicion of abuse; suspicious but not reported; and suspicious of abuse and reported) of 111 injury visits. Two techniques were used to validate the PHCPs' initial decision: expert review and provider retrospective self-assessment. Five child abuse experts reviewed clinical vignettes created by using data prospectively collected by PHCPs about the patient encounter. The PHCPs' opinions 6 weeks and 6 months after the injury-related visits were elicited and analyzed. RESULTS: PHCPs and experts agreed about the suspicion of abuse in 81% of the cases of physical injury. PHCPs did not report 21% of injuries that experts would have reported. Compared with expert reviewers, PHCPs had a 68% sensitivity and 96% specificity in reporting child abuse. A PHCP's decision to report suspected child abuse to CPS did not reduce the frequency of primary care follow-up in the 6 months after the index visit. PCHPs received information from their state CPS in 70% of the reported cases. CONCLUSIONS: Child abuse experts and PHCPs are in general agreement concerning the assessment of suspected child physical abuse, although experts would have reported suspected abuse to CPS more frequently than the PHCPs. Future training should focus on clear guidance for better recognition of injuries that are suspicious for child abuse and state laws that mandate reporting.