Objective: To assess the effectiveness and feasibility of implementing the Centers for Disease Control and Prevention (CDC) screening-based guidelines for preventing early-onset group B streptococcal sepsis. Methods: We compared prevalence of early-onset group B streptococcal sepsis after institution of the CDC screening-based protocol (October 1, 1995 through August 31, 1999) with that of historical controls (January 1, 1992 through June 30, 1995). We reviewed medical records for a cohort of deliveries of at least 23 weeks' gestation (January 1, 1996 through December 31, 1996) for group B streptococcal colonization status, risk factors, and intrapartum antibiotic prophylaxis. Results: The prevalence of early-onset group B streptococcal sepsis was 1.16 per 1000 (36 of 31,133) live births before and 0.14 per 1000 (four of 28,733) live births after institution of the CDC protocol (P < .001). Maternal colonization was known for 95.3% of the 7168 women who delivered (January 1, 1996 through December 31, 1996) at or after 37 weeks' gestation. Of 2174 women who qualified for intrapartum antibiotic prophylaxis, 1871 (86.1%) received it before delivery. There was 93.8% compliance with intrapartum antibiotic prophylaxis for women who delivered vaginally and 53.2% compliance for women who delivered by cesarean. Conclusion: Institution of the CDC screening-based protocol was accomplished at a specialty women's hospital, staffed by full-time faculty and community physicians, with 93.8% compliance for vaginal deliveries, and was associated with an 88% reduction in early-onset group B streptococcal sepsis. Copyright (C) 2000 The American College of Obstetricians and Gynecologists.