Eighteen- to 60-month-old iron-deficient anemic children given iron therapy (n=25) and a control group matched for mother's educational level showed no significant difference in mean mental development score at baseline. The control group's mean score was increased significantly over baseline score at 3 and 6 months and was significantly higher than the experimental group's mean score at 3 months. Although the experimental group demonstrated hematologic correction over 6 months, mean mental development score showed no significant improvement. Scores for an iron deficient not anemic group given iron (n=22), despite complete hematologic correction over the six months of observation, and for its control group, did not change significantly. Baseline scores for an iron-deficient not anemic placebo group (n=23) and for its control group were not significantly different. At 3 months the control group score had increased significantly, whereas that for the experimental group had not. When experimental and control subjects were matched on baseline mental development score, the control subjects experienced increases in scores over time, further confirming an impaired ability to improve scores with repeated testing in the experimental groups. Behavioral rating data (responsiveness to examiner, responsiveness to environment, and emotional tone) revealed significant differences between the iron-deficient anemic group and its control group at 3 and 6 months, with the control group rated more responsive, suggesting that iron deficiency, alone or in association with anemia, may have some lasting effect on behavior and development. Group differences were also found between the mean number of occurrences of multiple stressful events. Failure to show improvement in scores in the iron testable than were children in the control group, despite repeated testings, a theory supported by the infant behavior rating data. This may be related to some irreversible behavioral deficit or to an adverse environmental milieu (e.g., stress).