Neonates are among the most vulnerable patient groups for healthcare-associated infection with multiple endogenous and exogenous risks. Interpretation of neonatal bloodstream infection (BSI) rates requires stratification for case-mix. We assessed 1367 consecutive admissions to a single neonatal unit over a 34-month period. Four intrinsic and seven extrinsic risks were evaluated using Poisson regression analyses both individually and in combination. Nine of the 11 evaluated risk factors were significantly associated with BSI on univariate analyses. The only significant independent risks were parenteral nutrition, whether administered centrally or peripherally [incidence rate ratio (IRR): 14.2; 95% confidence interval (CI): 8.8-22.9; P<0.001], and gestational age <26 weeks (IRR: 2.5; 95% CI: 1.7-3.8; P<0.001). The rate of BSI per 1000 patient-days was 40 times higher in infants with both of these than in infants with neither. If validated in other settings, stratification of neonatal BSI rate by two unambiguous risk factors, parenteral nutrition and birth gestational age <26 weeks, offers a simple method to make meaningful intra- and inter-hospital comparisons.