The feasibility of using isotopic techniques to study Mg absorption and metabolism was explored in three full-term human infants.25Mg (98.8 atom %) was administered orally as an in vivo tracer. Fractional25Mg absorption, isotope retention, endogenous fecal Mg losses, and apparent Mg exchangeable pool size were then determined under three conditions of isotope administration:/) 20 mg25Mg, with single feeding; 2) 20 mg25Mg, distributed over a 24-h period; and 3) 60 mg25Mg, over a 24-h period. Mg isotope ratios were determined by inductively coupled plasma mass spectrometry. Fractional absorption was increased in all three infants after distributed versus bolus administration at the 20 mg dose; mean (±SD) fractional absorption was 64.0 ± 3.9 versus 54.3 ± 5.9%, respectively. 25Mg retention was also more in all three infants after distributed administration (55.8 ± 3.0 versus 44.3 ± 1.3% of dose). At the 60-mg25Mg dose, compared to 20 mg, fractional absorption was reduced but absolute isotope absorption more than doubled in all infants; urine isotope losses represented a similar fraction of the absorbed dose, thus,25Mg retention also more than doubled. Compared to the results of the isotope studies, net Mg absorption and balance were uninfluenced by total Mg intake. Isotope retention with distributed isotope administration resulted in measurable isotopic enrichment of plasma and erythrocytes at 72 h (i.e. plasma isotope enrichment was 6.3-10.2 and 19.2-23.5% for the 20- and 60-mg dose, respectively). With these doses, apparent Mg exchangeable pool size ranged from 5.5 to 7.6 mmol/kg body wt; these values showed a decrease with age both within and between infants. Our results indicate that Mg stable isotope studies may offer sufficient accuracy and reproducibility to permit meaningful investigations of Mg bioavailability and developmental changes in Mg metabolism in human infants.