OBJECTIVE: Characterize C-reactive protein (CRP) within 72 postnatal hours in early-onset sepsis (EOS). STUDY DESIGN: Secondary analysis of a prospective surveillance study of neonates with EOS 2015-2017. We examined CRP use by center and neonatal characteristics, and CRP levels by time, neonatal characteristics, clinical signs, and pathogen. RESULTS: CRP was obtained for 96/235 neonates with EOS, which varied by center (p < 0.001). 71/95 had CRP > 10 mg/L (1 missing). Neonatal characteristics with and without CRP did not differ. There was no relationship between CRP level and timing (p = 0.41) or neonate characteristics. Median CRP was higher with ≥5 vs <5 clinical signs (56, 23 mg/L; p = 0.002), and was not different in Gram-positive vs Gram-negative sepsis (43, 51 mg/L; p = 0.37) or preterm neonates who died vs survived (38, 28 mg/L; p = 0.37). CONCLUSIONS: Among neonates with EOS, CRP use varied by center. CRP levels did not differ by time, neonate characteristics, pathogen, or death. CLINICAL TRIALS REGISTRATION: ClinicalTrials.gov ID Early-Onset Sepsis an NICHD/CDC Surveillance Study (EOSII): NCT02410486.