Thirty Negro children with pica were compared nutritionally with 28 control children from the same socioeconomic background. The following findings were statistically significant: lower hemoglobin concentrations, lower levels of ascorbic acid in plasma, less adequate diets, more reported respiratory illness, more recorded days of hospitalization, and more recorded anemia in infancy. No difference was detected between the two groups in clinical evaluation of nutrition or in the number of physical defects. Evidence from anthropometry showed no inferiority in the pica group. Ancillary information indicated that many unfavorable familial and environmental circumstances were also associated with pica, as well as certain behavior disorders of childhood, especially those involving orality. In a double-blind experiment 16 children with pica received suitable doses of iron given intramuscularly, and 16 received similar injections of saline solution. Iron was not significantly more effective than saline solution in curing or improving the habit of pica. There was no correlation between changes in pica and changes in hemoglobin concentration. At 2 to 3 months after the treatment period, 10 treated children and 8 controls were without pica. The improvement in hemoglobin concentration was significantly greater in the treated group, but improvement in pica was approximately the same and highly significant in both groups. At 13 to 27 months after the treatment period, 11 of 14 children remaining in the treated group and 9 of 13 remaining controls were reported to be without pica. There were five severe relapses, two in the treated group and three in the controls, all of which took place in children with initial levels of hemoglobin above 10.5 gm/100 ml. These five children also came from the most seriously disturbed family situations in the whole series. If iron medication is a factor in curing certain cases of pica, the number of such cases in the present series is not sufficient to affect the statistical results.