In a cohort of term infants (n=91), followed from birth to 12 months, iron intake was examined by 24-h food records, and iron status by blood samples (haemoglobin (Hb)), mean corpuscular volume (MCV), serum values for iron, ferritin and transferrin, and erythrocyte protoporphyrin) at 2, 6 and 9 months. At 9 months of age, 5% had anaemia (Hb<105 g/l), but none had developed iron deficiency according to strict definitions used in this study (serum ferritin < 13 micrograms/l and transferrin saturation < 10%). Infants with high serum ferritin, serum transferrin and erythrocyte protoporphyrin values at one blood sampling also had high values at the following sample (tracking, r=0.45-0.80), suggesting that iron stores at delivery are an important determinant of iron stores during late infancy. Factors related to changes in serum ferritin were investigated by multiple linear regression. From 2 to 6 months, serum ferritin was negatively associated with knee-heel growth velocity (p=0.006) and positively with intake of infant formula (p=0.04). From 6 to 9 months it was negatively associated with intake of bread (p=0.001), and there was a trend for a positive association with intake of meat (p=0.07) and fish (p=0.08) and for a negative association with intake of cow's milk (p=0.07). In conclusion, those with a high growth velocity and a dietary pattern with a high intake of bread and a low intake of meat and fish had lower ferritin values and thereby an increased risk of depleting their iron stores later during infancy.