Background Iron deficiency almost invariably occurs in PI without supplementation. The AAP recommends an iron intake of 2-3mg/kg/day after hospital discharge. Despite the possible risks of excess oral iron, little data is available on the efficacy of intermediate levels of supplementation. Objectives to assess the INS of PI receiving about half the recommended level of iron after hospital discharge.Methods PI (BWt<1750g, GA<244d) were randomised to 3 groups at hospital discharge:- GpA received a lowbirthweight formula(80kcal/dl,0.9mg/dl iron,2.2g/dl protein) to 6mns corrected age; GpB, a term formula (66kcal/dl,0.5mg/dl iron,1.4g/dl protein); & GpC, the LBW formula until term, then the term formula. Growth, Hb & zinc protoporyrin (ZPP) were measured at discharge, every 2 weeks until term, then monthly. Formula intake was measured from difference in weight of preweighed cartoons of formula. Plasma ferritin (PF) was measured in a subgroup of infants (GpA & B only). Results 95 infants (BWt 1407±315, GA 219±16d) have been recruited and 71 followed to 6mns. Iron intake varied significantly between the groups (A 1.3±0.4mg/kg/day > C 1.0±0.4 > B 0.8±0.3, p<.0001), and was about half that recommended by the AAP. Only 2 infants (A=1, B=1) had Hb below the lower limit of normal for term infants of the same post-conceptional age (PCA). Subjects compared well with reference data for PI of the same post-natal age (PNA). Two infants had PF<10 mcg/l, neither were anemic (A=1, B=1). ZPP fell steadily from discharge to 6mns; and was significantly related to PNA (p<.001) but not PCA or PF. Overall Hb was higher in GpA than GpB or GpC (mean diff = 0.25g/dl, p=0.015). This did not appear to be due to differences in iron intake, as PF was higher in GpB than GpA (p=.04). We speculate that the difference in Hb was due to differences in protein intake(A=3.0±1.1mg/kg/d, B=2.7±0.9, C=2.7±1.0; A>B&C p<.0001), rather than iron intake. Conclusions (1) Iron intakes of 0.8-1.3 mg/kg/day maintained normal INS in PI, (2) ZPP is not a useful measure of INS in this population.