Background Neonatal sepsis is a leading cause of neonatal morbidityand mortality. There are no pathognomonic signs or symptomsto diagnose neonatal sepsis. Blood culture is the standard tool forsepsis diagnosis, but it is not available in most district hospitals. In2004, the Indonesian Pediatric Society (IPS) developed a medicalstandard of care to diagnose neonatal sepsis, but its accuracy hasnot been adequately verified.Objective To evaluate the diagnostic accuracy of the IPS medicalstandard of care 2004 to diagnose neonatal sepsis.Methods We conducted diagnostic testing at the PerinatalWard, Dr. Sardjito Hospital, Yogyakarta, from June to November2010. Inclusion criteria were neonates with signs and symptomsof infection. We excluded neonates with congenital anomalies,blood disorders, or whose mothers received peripartum antibiotictreatment. All neonates were assessed by the 2004 IPS medicalstandard of care for neonatal sepsis and by blood cultureexaminations.Results A total of 193 neonates with signs and symptoms ofinfection were evaluated. The IPS medical standard had asensitivity of 88% (95% CI 81 to 94), specificity of 17% (95%CI 2 to 25), positive predictive value of 53% (95% CI 45 to 60),negative predictive value of 57% (95% CI 39 to 75), positivelikelihood ratio of 1.06 (95% CI 0.94 to 1.19), and negativelikelihood ratio of0.71 (95% CI 0.36 to 1.42).Conclusion The 2004 IPS medical standard of care showedadequate sensitivity for diagnosing neonatal sepsis, but its lowspecificity limits its use as a diagnostic tool. [Paediatr lndones.2012;52:86-90].