BACKGROUND OF THE STUDYPreterm infants are managed with antibiotics for sepsis, including suspected or probable sepsis. This leads to a delayed and abnormal colonization of the gut with potentially pathogenic organisms and a microbiome, which lacks biodiversity and increases the risk for late-onset sepsis (LOS). Probiotics have been proven to reduce the risk for necrotizing enterocolitis, but evidence for prevention of LOS is inconclusive. Probiotic effect depends also on the strain used, dose and indication for use. This study evaluated Bacillus clausii probiotic administered prophylactically to preterm neonates for prevention of LOS.OBJECTIVESTo study B.clausii given prophylactically to preterm neonates for prevention of LOS.DESIGNDouble-blinded, placebo-controlled, randomized trial.SETTINGSTertiary care neonatal unit in India.PARTICIPANTSConsecutive preterm neonates <34 weeks, admitted from 1 March 2012 to 28 February 2014 were stratified as extreme preterm and very preterm.INTERVENTIONRandomized to receive either probiotic or placebo for 6 weeks, discharge from hospital, death or occurrence of sepsis, whichever was earlier.PRIMARY OUTCOMEIncidence of definite and probable LOS in probiotic group compared with placebo.RESULTSOf 326 eligible preterm infants, 244 were enrolled and 82 were excluded. Of these, 120 were stratified as extreme preterm and randomized to receive placebo (n = 59) and probiotic (n = 61). Of 124 babies stratified as very preterm, an equal number was randomized to receive placebo (n = 62) and probiotic (n = 62). There was no significant difference in the incidence of LOS between the two arms in the extreme preterm group [29% vs. 23%; relative risk (RR) 1.27; 95% confidence interval (CI) 0.88-1.66; p = 0.36] and the very preterm group (13% vs. 10%; RR 1.33; 95% CI 0.96-1.70; p = 0.32). Full feeds were achieved significantly faster in the probiotic group in both the extreme preterm (RR 0.82; 95% CI 0.74-0.88) and the very preterm (RR 0.67; 95% 0.32-0.77).CONCLUSIONSProphylactic administration of B.clausii to preterm neonates did not result in a significant difference in the incidence of LOS as compared with placebo.