Research about neonatal outcomes among late preterm infants (34 weeks through 36 6/7 weeks of gestation) is limited. Understanding which late preterm infants are at risk for neonatal morbidity or mortality is necessary to improve health outcomes and reduce hospital costs. We conducted a population-based cohort study of "healthy," singleton late preterm infants vaginally delivered in Massachusetts hospitals to Massachusetts residents between 1998 and 2002. We compared the incidence of neonatal morbidity (postdelivery inpatient readmissions, observational stays, or mortality) between "healthy," late preterm infants with and without infant, obstetric, and sociodemographic factors by calculating risk ratios adjusted for confounding. Of the 9552 late preterm, "healthy" infants, 4.8% had an inpatient readmission and 1.3% had an observational stay. Infants with neonatal morbidity were more likely to be firstborn, be breastfed at discharge, have labor and delivery complications, be a recipient of a public payer source at delivery, or have an Asian/Pacific Islander mother. Non-Hispanic blacks had a decreased risk for neonatal morbidity compared to other racial/ethnic groups. Knowledge of risk factors for neonatal morbidity among "healthy" late preterm infants can be used to identify infants needing closer monitoring and earlier follow-up after hospital discharge.