The objective of this study was to evaluate the benefit of screening for anemia in infants in relation to their previous diet. The iron status of 854 nine-month-old infants on three different feeding regimens and on a regimen including iron dextran injection was determined by analysis of hemoglobin, serum ferritin, and erythrocyte protoporphyrin levels and of serum transferrin saturation. Infants were categorized as having iron deficiency if two or three of the three biochemical test results were abnormal and as having iron deficiency anemia if, in addition, the hemoglobin level was less than 110 gm/L. The prevalence of iron deficiency was highest in infants fed cow milk formula without added iron (37.5%), intermediate in the group fed human milk (26.5%), much lower in those fed cow milk formula with added iron (8.0%), and virtually absent in those injected with iron dextran (1.3%). The corresponding values for iron deficiency anemia were 20.2%, 14.7%, 0.6%, and 0%, respectively. The use of iron supplements is therefore justified in infants fed cow milk formula without added iron, even when there is no biochemical evidence of iron deficiency. The low prevalence of iron deficiency in the group fed iron-fortified formula appears to make it unnecessary to screen routinely for anemia in such infants. These results also support the recommendation that infants who are exclusively fed human milk for 9 months need an additional source of iron after about 6 months of age.