Background Sepsis is a major problem in neonatology with high mortality (13–25%). Early diagnosis leads to good results.

Materials We studied 110 newborns with sepsis. We divided them into 2 groups. 1st group, 60 full term newborns. 2nd group, 50 preterm newborns.

Methods The study was an open prospective research, approved by the clinical bioethics commission. We studied C-reactive protein (CRP) by the latex-agglutination method, and procalcitonin (PCT) by the immunoluminometric method. Detection of bacteria in blood was done with polymerase chain reaction (PCR) and by traditional culture methods.

Results During sepsis the level of CRP in full term newborns increased (96–192 mg/l; norm<6 mg/l) but the level was highest at 36 hours (measured every 12 hours). During sepsis, in preterm newborns the level of CRP was normal in 93% and increased in 7%. After 3 hours of developing sepsis, the level of PCT increased in both groups (PCT >2 ng/ml; norm<0.05 ng/ml), but determination of procalcitonin was limited. Research was statistically reliable (P<0.05).

Conclusions 1) PCT may be indicated as a useful surveillance marker in newborns at risk for severe nosocomial infections and in those presenting with clinical signs of sepsis. 2) CRP and PCT are reliable diagnostic markers in neonatal sepsis, among full term newborns, from 2 days of life, but PCT was more informative, quick and correct than CRP. 3) CRP is not a reliable diagnostic marker for preterm newborns in neonatal sepsis.