The objective of this prospective, cohort study was to compare the nutritional status of full-term infants who were fed human milk (BF, n = 29), formula (FF, n = 30) or evaporated milk formulae (EM, n = 30) for at least 3 months. Infants were seen at enrollment, 3 and 6 months, at which times a blood sample, diet record and anthropometric data were collected. Infants in the EM group received solids earlier (12 +/- 5 weeks) than did FF infants (15 +/- 4 weeks), and both were earlier than BF infants (19 +/- 4 weeks). Only 26% of the EM fed group received iron supplements as ferrous sulphate drops. Seven BF, 12 FF and 20 EM had abnormal ferritin values (< 10 ng ml-1) at 6 months. Copper intake was lower in the EM infants at 3 and 6 months. However, plasma copper and erythrocyte copper zinc superoxide dismutase (ZnCuSOD) levels did not differ between groups. Selenium intake was lower in the EM group (5 +/- 1 and 10 +/- 5 micrograms d-1; 3 and 6 months) than in the FF infants (13 +/- 4 and 19 +/- 7 micrograms d-1; 3 and 6 months). Erythrocyte SeGHSPx levels in EM infants were lower at 6 months (EM, 33.2 +/- 3.4; FF. 35.2 +/- 3.9: BF, 36.1 +/- 3.8 mU mg Hb-1). Thiamin intake (0.99 +/- 0.08 and 1.24 +/- 0.32; 3 and 6 months, mg 1000 kcal-1) was higher in the FF group than in EM infants (0.38 +/- 0.39 and 0.66 +/- 0.38; 3 and 6 months). There were more (13%) abnormal thiamin assays in the EM group at 6 months than in the BF and FF infants (0%). In conclusion, infants fed evaporated milk formula receive adequate copper but may not receive enough thiamin or selenium. Unless supplemented from birth with medicinal iron, intakes of iron will be inadequate.