Because of reluctance to use radioisotopes for studies of iron absorption in children, we have explored the feasibility of using the least abundant stable isotope of iron, 58Fe (natural abundance, 0.322 weight %) in a study of nonheme iron absorption. With a balanced cross-over design, each of 16 school-age children was fed a standardized lunch on 3 consecutive days and, 28 days later, an alternate standardized lunch on 3 consecutive days. The lunch included either a beef patty or a beef-soy patty. The mass isotope ratio, 58Fe/57Fe (MIR58/57), was measured in blood by inductively coupled plasma mass spectroscopy before and 14 days after (i.e. study day 15) consuming the three lunches. The MIR58/57 on study day 15 was used as a baseline value for lunches fed on study days 29, 30, and 31. Incorporation of 58Fe into erythrocytes was greater from the lunch with beef patty than from the lunch with beef-soy patty (geometric mean values 2.02 and 1.05% of the dose, p less than 0.03). Based on the similarity of our results with those obtained in adults with radioisotopes, we conclude that 58Fe is a satisfactory tag for studies of nonheme iron absorption from meals.