The possibility of reducing calcium inhibition of iron absorption by decreasing calcium intake in lunch and dinner meals, which provided the most dietary iron, was examined in 21 healthy female volunteers. During a 10-d period, nonheme iron in all meals was extrinsically labeled with radioisotopic iron to a uniform specific activity. Iron absorption from two identical 10-d periods was compared when meals were labeled with two different iron radioisotopes and when the same amount of calcium (937 mg) was distributed in two ways, in either mainly breakfast and late evening meals or more evenly in all meals. About 30-50% more iron was absorbed when no milk or cheese was served with lunch or dinner. The difference was statistically significant. Median iron requirements (1.61 mg/d) calculated from body weight and menstrual iron losses agreed with the mean value of median iron absorption in the two 10-d periods (1.54 mg/d), which supports the validity of the present method. A reasonable separation of calcium and iron intakes would improve iron nutrition.