The aim of this study is to evaluate the effect of administration of recombinant human granulocyte colony-stimulating factor (rhG-CSF) in newborn infants with the diagnosis of neonatal sepsis in a randomized, controlled fashion.Thirty nine preterm neonates with neonatal sepsis admitted to the Neonatal Intensive Care Units of Cairo University Hospital were studied. Twenty neonates were randomized to receive rhG-CSF in a dose of 10ug/kg/day subcutaneously in addition to conventional therapy. Nineteen neonates were randomized not to receive rhG-CSF. Complete blood counts with manual differential counts, sepsis scores, Score for neonatal acute physiology (SNAP) were obtained on study entry and on study days one two and three thereafter. Morbidity and mortality were recorded as secondary outcome measures. We found that the administration of Recombinant Human Granulocyte Colony Stimulating Factor to premature neonates with sepsis in a dose of 10ug/kg/day subcutaneously on 3 subsequent days compared to a control group not receiving G-CSF did not significantly alter the circulating absolute neutrophil count, significantly increased the I/T ratio and significantly increased the immature neutrophil count. There was also a significant increase in the absolute monocyte count. No significant change in absolute lymphocyte count, total white blood cell count, platelet count, hemoglobin concentration and hematocrite were noted. There was also no significant difference in severity of acute illness as determined by SNAP, and no significant difference in morbidity or mortality between the 2 studied groups. The drug was tolerated well and administration was not associated with any observed side effects.