INTRODUCTION: Two strategies are available for prevention of early-onset group B streptococcal (GBS) sepsis - clinical risk factor-based screening and routine culture-based screening of pregnant women for GBS colonisation. In our hospital, we switched from the former to the latter approach in 2014. METHODS: We compared the incidence of early-onset GBS sepsis during 2001-2015 between infants born to pregnant women who were screened for GBS colonisation and those born to women who were not screened. RESULTS: Among 41,143 live births, there were nine cases of early-onset GBS sepsis. All infants with GBS sepsis were born to pregnant women who were not screened for GBS colonisation. The incidence of early-onset GBS sepsis among infants of women who were not screened was 0.41 per 1,000 live births (95% confidence interval [CI] 0.19-0.77) when compared to infants of women who were screened, for whom the sepsis incidence was zero per 1,000 live births (95% CI 0-0.19; p = 0.005). CONCLUSION Our data suggests that routine culture-based screening of pregnant women for GBS colonisation is a better preventive strategy for early-onset GBS sepsis in neonates when compared to clinical risk factor-based screening.