Abstract: BACKGROUND: Iron (Fe) is essential for growth, but optimal intake is controversial. Our NICU practice was to supplement 2 mg/kg/d Fe for all preterm infants receiving human milk when they achieved full feeding volume. Adjusting Fe supplementation based on ferritin levels is thought to better address physiologic requirements. Our objective was to assess the impact of therapeutic monitoring of ferritin levels on the initiation and dosing of iron supplementation, hematocrit, transfusions, and oxygen radical diseases in preterm infants. METHODS: Preterm infants (< 32 weeks gestation, n = 100) were included. Ferritin was measured when full feeds were achieved, and then every 2 weeks. Fe was started at 2 mg/kg/d or continued at current dose for ferritin 40–300μg/L, increased by 1–2 mg/kg/d for < 40μg/L, or discontinued for > 300μg/L. Outcomes were compared with a historical control group. RESULTS: Ferritin levels were not predictable by dietary or transfusion histories. Using the ferritin protocol, 70% of infants received Fe at the time of full feeds, compared to 100% of controls. In contrast, all infants received Fe 4 weeks later, compared to 87% of controls. Mean age at Fe initiation increased (14.8±6.3 to 21.0±11.76 days). Peak doses were higher, with 32% receiving > 2 mg/kg day by 6 weeks, with fewer transfusions. The incidence of bronchopulmonary dysplasia and necrotizing enterocolitis did not change. CONCLUSION: An iron protocol based on ferritin levels results in later initiation, higher doses, and fewer transfusions, without increasing oxygen radical diseases.