Background: Infants at risk for early onset sepsis (EOS) are evaluated based on current CDC/AAP guidelines (newborn nursery protocol). However, many well-appearing infants receive unnecessary empiric antibiotics despite extremely low incidence of EOS. Quantitative risk-based prediction models such as Kaiser Permanente's Sepsis Risk Calculator (SRC)®, is an effort to minimize antibiotic exposure in this cohort. It is not certain however,that SRC will function equally well among high-risk population of infants whose mothers have a high rate of limited to no prenatal care (PNC). Objective: To assess the rate of antibiotic usage and laboratory work-up in infants evaluated for EOS by the current newborn nursery (NBN) protocol and compare them with the rates of antibiotic usage and laboratory tests, if the same infants were to have been evaluated by SRC. Design/Method: Retrospective review of infants admitted to NBN ≥35 wks gestational age (GA) between June, 1st, 2016 and May, 31st, 2017 who had blood culture drawn during hospital stay. Maternal and infant data were obtained from electronic medical records and put in the on- line SRC to obtain risk estimate and recommendations for treatment. The values were then compared with antibiotic usage rates in these infants that had been treated based on the existing NBN protocol. Results: 175 out of 208 infants were included. 33 infants met exclusion criteria (direct NICU admits, congenital anomalies or infants whose maternal details could not be obtained). The median GA was 37 wks (IQR 36,41) with a mean BW of 3047±572 grams; 47% were females. 19% mothers had limited/no PNC; 32% were GBS positive and 35% were GBS unknown. Chorioamnionitis was diagnosed in 28% of mothers and UTI in 3%. Antibiotics were recommended in only 23/175 (13%) infants by SRC as compared to 98/175 (56%) infants who actually received antibiotics based on the NBN protocol (p<0.0001). Laboratory work-up was recommended in 19% of infants by SRC . No difference was noted in the WBC count (p=0.556) and I:T ratio (p=0.18) between infants with positive and negative blood cultures. 4 infants had positive blood cultures, all with exposure to maternal chorioamnionitis that were started on empiric antibiotics based on the CDC/AAP guidelines. Two of these infants qualified for antibiotics by SRC. One organism was considered a contaminant. Only one clinically well appearing infant, treated for E.coli bacteremia, did not qualify for antibiotics by SRC. Conclusion: The application of SRC would have realized a significant reduction in the rate of antibiotic usage and lab work-up in infants treated for EOC as compared to the current guidelines. SRC would have led to the treatment of 2 out of 3 culture positive infants treated for EOS. Prospective trials evaluating the safety of SRC in high risk population are warranted.