OBJECTIVE: To evaluate how UK maternity unit protocols conformed to national recommendations for preventing early-onset neonatal Group B Streptococcal (GBS) disease. METHODS: In December 2005, all UK obstetric maternity units were contacted and asked to provide a copy of their protocol on preventing GBS disease. Information was extracted on the protocol's recommendations, its development date and the evidence cited. The protocol's recommendations were then compared against the recommendations in the Royal College of Obstetricians and Gynaecologists (RCOG) guideline. RESULTS: Protocols were obtained for 171 of the 227 units (75%), of which 120 were developed after the guideline has been published. There were 134 protocols (78%) that followed the RCOG prevention strategy, recommending a risk-based approach to selecting women for intrapartum antibiotic prophylaxis (IAP). However, the sets of risk factors named as indications for IAP differed between the protocols and only 34 of these 134 protocols were entirely consistent with the guideline. The 37 protocols (22%) that did not follow the RCOG prevention strategy recommended IAP for some risk factors but only if a bacteriological test was also GBS positive. CONCLUSIONS: There are considerable differences in the GBS protocols used in maternity units in the UK despite the availability of a national guideline. Consequently, some high-risk women may not receive IAP while some women without risk factors are treated needlessly. While local adaptation may be for legitimate reasons, the processes used in some units seem to require improvement.