Data are scarce regarding mineral bioavailability from human milk in older infants who may also be receiving solid foods (beikost). We measured the absorption of Ca, Zn, and Fe in 14 healthy, nonanemic 5-7-mo-old breast-fed infants whose mothers milk was extrinsically labeled with stable isotopes (44Ca, 70Zn, and 58Fe) of these minerals. In addition, Ca and Zn stable isotopes (46Ca and 67Zn) were administered i.v., and a second isotope of Fe (57Fe) was given orally without food as a non-meal dose. Subjects were not receiving any artificial infant formula or cow's milk, but most (10/14) were receiving beikost. Ca and Zn absorption was calculated using the urinary excretion of the isotopes during the 24 h after dosing (Ca) or their urinary ratio 72 h after dosing (Zn). Fe absorption was calculated using the red blood cell incorporation at 14 d. Fe absorption averaged 20.7 +/- 14.8% from the 58Fe given with human milk (geometric mean, 14.8%) and 17.7 +/- 15.1% (geometric mean, 11.0%) from the 57Fe non-meal dose. Ca absorption averaged 61.3 +/- 22.7% and Zn absorption (n = 10) averaged 49.5 +/- 18.5%. Absorption of Fe (natural logarithm) from the non-meal Fe dose (57Fe) but not from the human milk (58Fe) was significantly negatively correlated to serum ferritin (r = -0.70, p = 0.007 versus r = -0.35, p = 0.24). At the intake levels in this study, total daily Fe, Ca, and Zn intakes from beikost were not significantly correlated to their fractional absorption from breast milk, but Fe intake from beikost was significantly negatively correlated to absorption of Fe from the non-meal dose (r = -0.61, p = 0.021). We conclude that minerals are well absorbed from human milk in older infants after the introduction of beikost to the diet.