Four behavioral rating scales (BRS) (CHEOPS, CHIPPS, FLACC and OPS) assessing postoperative pain in children aged 1-7 years were studied to compare their psychometric properties, sensitivity and specificity. One hundred and fifty children included in this prospective longitudinal study were videotaped to analyze retrospectively peri-operative behaviors. Pain and anxiety were evaluated by children or by their parents prospectively. At the end of the study, four observers rated the peri-operative videos using the four BRS. Because self-reporting cannot be used for all the children, facial expression of pain was analyzed from the videos to create a Facial Action Summary Score (FASS) which was considered as a reference for the study of validity of the four BRS. Internal validities were excellent but external validities were mixed. The FLACC seems to be better adapted to assess post-surgical pain in children between 1 and 7 years old. Nevertheless, it was significantly correlated with anxiety measures. Moreover, the analysis of sensitivity and specificity using both self-reporting of pain and FASS showed that some children were still under-evaluated. The multivariate analysis underlines silence as a high risk factor of misevaluating postoperative pain. In conclusion, this study highlights the difficulty of discriminating pain intensity from anxiety when using the four BRS and that postoperatively, nearly one child in 10 was misevaluated.