Accumulation of bile acids (BAs) may mediate development of necrotizing enterocolitis (NEC). Serial fecal samples were collected from premature infants with birth weight (BW) ≤ 1800 g, estimated gestational age (EGA) ≤ 32 weeks, and <30 days old prior to initiation of enteral feeding. Nine infants that developed Bell's Stage ≥ II NEC were matched with control infants based on BW, EGA, day of life (DOL) enteral feeding was initiated and DOL of the first sample. From each subject, five samples matched by DOL collected were analyzed for BA levels and composition. Fifteen individual BA species were measured via LC-MS/MS and total BA levels were measured using the Diazyme Total Bile Acid Assay kit. No statistically significant differences in composition were observed between control and NEC at the level of individual species (p = 0.1133) or grouped BAs (p = 0.0742). However, there was a statistically significant difference (p = 0.000012) in the mean coefficient of variation (CV) between the two groups with infants developing NEC having more than four-fold higher mean CV than controls. Importantly, these variations occurred prior to NEC diagnosis. These data suggest fluctuations in total fecal BA levels could provide the basis for the first predictive clinical test for NEC.protective responses 20 , and the concomitant dysregulation of hepatic BA transporters 21 during development of NEC. This novel paradigm encompasses major disease-associated findings and risk factors: 1) the more premature an infant, the more likely they will develop NEC 10,11 . Many of the key processes of BA homeostasis are not fully developed in neonates, but reach maturity at weaning [22][23][24][25][26][27] . 2) Formula-fed preemies are 6-10 times more likely to develop NEC 12 and have higher fecal BA levels than breast-fed preemies 28 . Further, formula feeding is required to develop experimental NEC 29,30 . 3) The majority of BA reclamation occurs in the ileum and colon, the primary sites of NEC injury 31 , and ileal BA levels increase 24-hours before inflammation or histological damage in experimental NEC 32 . 4) Bacterial colonization is required for both development of NEC and formation of secondary BAs. While no specific pathogen has been conclusively associated with NEC 33-43 , disease cannot be developed in germ free conditions 13,14 and pneumatosis intestinalis, the most pathognomonic radiological finding of NEC, is thought to be caused by bacterial overgrowth 44 . Bacteria are also required for formation of more toxic, hydrophobic BAs 45-47 , which are higher in both experimental 18 and human NEC 48 .To further characterize how BAs are dysregulated in human NEC, we prospectively collected fecal samples from premature infants and determined BA levels and composition from matched-pairs diagnosed with NEC versus those without NEC. Contrary to a previous study 48 , we found no statistically significant differences in BA composition between control and NEC. However, there was a statistically significant difference (p = 0.000012) in the mean...