Background: Necrotizing enterocolitis (NEC) predominantly affects preterm infants and is associated with high mortality and morbidity. There is no composite scoring tool that predicts the severity of NEC in preterm infants early in the disease process before it is established. Methods: A retrospective chart review was performed on NEC cases (n = 41) defined as infants who developed NEC stage ≥ 2 with pneumatosis intestinalis, and controls (n = 41) defined as infants with NEC stage 1, to develop a scoring tool that utilized the preterm infant's clinical status, feeding practices, radiographic and laboratory findings at set time-points (7 days, 2 days and 24 and 12 hours) before the suspicion of NEC to predict the severity of NEC. Repeated measures analysis along with sensitivity, specificity and area under curve were measured to compare the composite risk score between cases and controls. Results: The composite risk score was higher among the cases at 24 hours (2.73 ± 0.20 vs. 2.20 ± 0.20, p < .05) and 12 hours (4.20 ± 0.22 vs. 2.50 ± 0.22, p < .001) but not on 2 day (2.37 ± 0.17 vs. 2.18 ± 0.17, p = .42) prior to the suspicion of NEC to predict the severity of the disease. Feeding practices, gastric residuals, and laboratory signs of inflammation were main determinants of the final score. The most discriminatory score to differentiate between NEC stage 1 and NEC stage 2 and 3 was ≤ 2. Conclusions: This scoring system is easy to use as it utilizes normally collected data and can predict the severity of the disease enabling the physician to make decisions regarding appropriate timing of correct treatment. Further larger prospective studies are required to validate this pilot study describing NEC severity prediction tool.