ABSTRACT. The Cu intake and status of 106 very low birth weight (VLBW) infants (birth weight 1152 ± 251 g, gestational age 29 ± 3 wk, mean ± SD) were determined approximately 1 mo before hospital discharge, at discharge (time 0), and at 3, 6, 9, and 12 mo ± 3 wk corrected for gestational age. Infants were fed either formula plus supplemental Zn/Cu (SUPPL, n = 29); formula plus placebo (PLAC, n = 26); or a low birth weight formula (LBWF, n = 26) or were breast-fed (BRMLK, n = 25).Plasma Cu levels in the formula-fed infants increased significantly at each time period with no significant differences between feeding groups. Hair Cu was significantly higher in the SUPPL group compared to the PLAC, LBWF, and BRMLK groups at 3 and 6 mo. Erythrocyte Cu,Zn-superoxide dismutase (CuZnSOD) activity was lowest in the PLAC group. Cu intake was positively correlated with both hair Cu (r = 0.291, p < 0.0001) and erythrocyte CuZnSOD activity (r = 0.281, p < 0.001) but not with plasma Cu. An increasing number of formula-fed infants had very low CuZnSOD activity (less than 2 SD below mean) with increasing age (n = 1,2,8, 11, and 13 infants at times 0, 3, 6, 9, and 12 mo, respectively). At 12 mo, approximately one third of the formula-fed VLBW infants in this study had low Cu status as assessed by CuZnSOD activity. Infants with the lowest CuZnSOD activity were those with the largest weight gains from 0 to 6 mo and were observed in all formula-fed groups. Thus, these data suggest that in formula-fed infants erythrocyte CuZnSOD activity is a more appropriate indicator of Cu status in the VLBW infant than plasma Cu levels and that the fastest growing VLBW infants may be at risk for Cu deficiency.