BACKGROUND: Whether lactoferrin (Lf) binds iron to facilitate its absorption or to sequester iron from potential enteropathogens remains uncertain. Bovine Lf is added to many infant formulas, but previous studies in infants reported that Lf had no effect on or inhibited iron absorption. The effects of the apo (iron-free) or the holo (iron-loaded) forms of Lf on iron absorption are unclear. OBJECTIVES: Our objective was to compare iron absorption from a maize-based porridge containing: 1) labeled ferrous sulfate (FeSO4) alone; 2) labeled FeSO4 given with bovine apo-Lf; and 3) intrinsically labeled bovine holo-Lf. METHODS: In a crossover study, we measured iron absorption in Kenyan infants (n = 25; mean ± SD age 4.2 ± 0.9 months; mean ± SD hemoglobin 109 ± 11 g/L) from maize-based test meals containing: 1) 1.5 mg of iron as 54Fe-labeled FeSO4; 2) 1.42 mg of iron as 58Fe-labeled FeSO4, given with 1.41 g apo-Lf (containing 0.08 mg iron); and 3) 1.41 g holo-Lf carrying 1.5 mg iron as 57Fe. The iron saturation levels of apo- and holo-Lf were 0.56% and 47.26%, respectively primary outcome was fractional iron absorption (FIA), assessed by erythrocyte incorporation of isotopic labels. RESULTS: The FIA from the meal containing apo-Lf + FeSO4 (geometric mean, 9.8%; -SD and +SD, 5.4% and 17.5%) was higher than from the meals containing FeSO4 (geometric mean, 6.3%; -SD and +SD, 3.2% and 12.6%; P = 0.002) or holo-Lf (geometric mean, 5.0%; -SD and +SD, 2.8% and 8.9%; P