In a previously published prospective study, we followed the development of allergic symptoms in term infants with a slightly reduced birthweight (− 1 SD to − 2 SD). These children received, according to local routing early feeding with cow milk formula in order to diminish such neonatal problems as hypoglycemia and hyperbilirubinemia. Of 216 infants 207 were observed for allergic symptoms up to 18 months of age. One group (F) received cow milk formula during the first days of life before the mother's breastmilk production started and was then breastfed; the other (B) was not given any formula before normal breastfeeding started. Unexpectedly, we found fewer allergic symptoms, in particular allergic skin problems, in the group fed cow milk, the difference being concentrated to children with a family fistory of allergic symptoms. At 5 Years of age 183 of the 207 children have been reinvestigated. Mild symptoms of allergy (suspected and obvious) were found in 22% (F) and 27% (B) respectively(NS). Moderate and severe symptoms of allergy (obvious) were found in 4.2% (F) and 4.5% (B). In the subgroup with a double family history of allergic symptoms, 28% (7/25, F) and 59% (10/17, B) had symptoms of allergy (p<0.05). This difference was even more pronounced when laboratory tests in favour of atopic diagnosis were included, 1.4% (F) and 53%(B) respectively (p< 0.05). Thus at 5 years we still find a significantly lower frequency of allergic symptoms in the subgroup fed cow milk formula early with a family history of allergic symptoms. This difference was mainly caused by a reduced incidence of mild symptoms, some of which might not be true allergic symptoms. The incidence of moderate to severe disease was not affected, which can possibly be due to the relatively small number of children, expecially in the high risk group. Thus feeding with a high allergenic load during the first days of life seems, in the long run, to be ineffective in inhibiting the development of Significant (moderate and severe) atopic disease.