Four strategies for prevention of early onset neonatal group B streptococcal (GBS) sepsis were considered: A: routine antenatal screening for GBS vaginal carriage at 26-28 weeks' gestation and intrapartum antibiotic prophylaxis for all carriers; B: screening as above and prophylaxis only for carriers with risk factors for sepsis; C: prophylaxis for all women with risk factors; D: as for C plus screening at 37 weeks' gestation and prophylaxis for carriers. The outcomes considered for each option were: the proportion of women given prophylaxis; the risk of anaphylaxis; cases of neonatal GBS sepsis and deaths prevented; costs of screening, prophylaxis and of acute care of remaining cases. Published local and overseas studies of neonatal GBS sepsis, effectiveness of antenatal screening and prophylaxis and estimated costs were evaluated. Any of the proposed strategies can prevent a significant proportion of cases of neonatal GBS sepsis and a strategy for prevention of neonatal group B streptococcal sepsis should be part of routine obstetric practice. Strategy C is simple, effective, inexpensive and avoids unnecessary antibiotic use; it is recommended particularly when antenatal care is provided mainly in community or private practice. Strategy A (using vaginal and rectal swabs for screening) could prevent more cases, but at greater cost which could be justified only if protocols can be properly implemented and monitored.