Although the rate of early onset sepsis in the near-term neonate is low (one to eight of 1,000 cases), the rate of mortality and morbidity is high. As a result, infants receive multiple, broad-spectrum antibiotic therapy, many for up to 7 days despite blood cultures showing no growth. Maternal intrapartum antibiotic prophylaxis and small blood volume collections from infants are cited as reasons for the lack of confidence in negative culture results. Incorporating an additional, more rapid test could facilitate a more timely diagnosis in these infants. To this end, a 16S rDNA polymerase chain reaction (PCR) assay was compared to blood culturing for use as a tool in evaluating early onset sepsis. Of 1,751 neonatal intensive care unit admissions that were screened, 1,233 near-term infants met inclusion criteria. Compared to culture, PCR demonstrated excellent analytical specificity (1,186 of 1,216, 97.5%) and negative predictive value (1,186 of 1,196, 99.2%); however, PCR failed to detect a significant number of culture-proven cases. These findings underscore the cautionary stance that should be taken at this time when considering the use of a molecular amplification test for diagnosing neonatal sepsis. The experience gained from this study illustrates the need for changes in sample collection and preparation techniques so as to improve analytical sensitivity of the assay.