Objective: The aim of this study was to evaluate the usefulness of maternal serum and vaginal fluid C-reactive protein (CRP) determinations in the prediction of neonatal congenital infection.Study Design: Fifty women between 24 and 36 weeks gestation, complicated by preterm premature rupture of membranes (pPROM), were divided into two groups according to the presence (n ¼ 14) or absence (n ¼ 36) of early-onset newborns' infection.Result: Maternal serum and vaginal fluid CRP concentrations were comparable between both groups. Serum CRP levels X10 and X15 mg l -1 predicted neonatal infection with a sensitivity of 47 and 47%, specificity of 63 and 76%, positive predictive value (PPV) of 38 and 47%, and negative predictive value (NPV) of 72 and 76%, respectively. The cutoff value of vaginal CRP X2.4 mg l -1 predicted infection with a sensitivity of 71%, specificity of 47%, PPV of 34%, and NPV of 81%. Receiver-operating characteristic curve analysis revealed that the predictive performance of CRP was poor.Conclusion: Maternal serum and vaginal fluid CRP determinations after pPROM are of poor predictive value in neonatal early-onset infection prediction.