Infants need arachidonic acid (AA; C20:4n-6) for eicosanoid synthesis and deposition in growing tissues, including brain. Human milk supplies preformed AA in amounts considered to meet accretion in membrane-rich tissues, but vegetable oil-based infant formulas do not contain AA. We studied two groups of ten healthy infants, each fed human milk or formula, and analyzed plasma lipid composition. Percentage contributions of AA to plasma phospholipids were stable over two months after birth in breast-fed infants, but infants fed formula developed significantly (P < 0.05) lower levels at the ages of two weeks (formula 6.9% vs. breast 9.4%, w/w), one month (6.2 vs. 9.1%), and two months (5.7 vs. 8.4%). In a second trial, we randomized infants to receive (from birth to age four months) formula without or with both AA and docosahexaenoic acid (DHA; C22:6n-3) at levels typical for mature human milk. Infants fed conventional formula showed a continuous decrease of phospholipid AA over time, whereas feeding of formula supplemented with AA and DHA led to significantly higher AA levels, similar to those in breast-fed infants (two months: supplemented 9.6% vs. unsupplemented 7.1%; four months: 8.7 vs. 6.6%). In order to estimate infantile capacity for endogenous synthesis of AA, we fed four term neonates with newly diagnosed phenylketonuria (mean age 18 d) a formula with all fat contributed by corn oil, which has a higher natural 13C-enrichment than European human milk or formula. Analysis of 13C-enrichment in plasma fatty acids over four days allowed us to estimate infantile AA synthesis. We found an increased 13C-value in plasma AA of all infants, which indicates that term neonates can synthesize AA. However, with a simplified isotope balance equation, we estimate that endogenous synthesis contributed only about 23% of total plasma arachidonic acid by day four. We conclude that full-term infants fed formula may require a dietary supply of some preformed AA if the biochemical status of breast-fed infants is to be achieved.