Erythrocyte superoxide dismutase (ESOD) activity reflects copper utilization and the risk of copper deficiency. To investigate the possible effects of inorganic iron on the metabolism of copper in low birth weight infants, we have measured ESOD activities in three groups of infants receiving different iron supplements. Fifty-five low birth weight infants were randomly assigned to receive daily from 28 d either 13.8 mg (HiFe), 7 mg (MidFe), or no elemental iron (NatFe) as iron edetate. At 27 d, 8, 12, and 20 wk postnatal age, infants were weighed and measured and hematologic indices, plasma ferritin, zinc, and copper concentrations, and ESOD activities were assayed. Anthropometrical and hematologic indices and plasma copper and zinc concentrations did not differ among treatment groups at any time, but at 20 wk, plasma ferritin concentrations [(micrograms/L) mean; SD] were lower in the NatFe group (17; 2.0) than in the HiFe group (32; 1.9: 95% confidence interval for mean difference 6.6 to 22.0, p less than 0.01). ESOD activities (U/g Hb) were similar in HiFe (1447; 263), MidFe (1552; 322), and NatFe (1538; 382) groups at 27 d, but by 20 wk activities in the HiFe group (1537; 211) were lower than in the MidFe (1789; 403: 95% confidence interval 38 to 466, p less than 0.05) and NatFe (1858; 304: 95% confidence interval 150 to 492, p less than 0.01) groups. The lower ESOD activities found in the HiFe group at 20 wk may reflect altered copper metabolism induced by the iron supplement, but the clinical importance of this observation is unknown.