Docosahexaenoic acid (DHA), the main n-3 polyunsaturated fatty acid (PUFA) in membranes, is particularly abundant in brain cells. Decreased cerebral concentrations of DHA, resulting from dietary n-3 deficiency, are associated with impaired cognitive function. Because the cellular causes of this impairment are still unknown, we need in vitro models that mimic the variations in n-3/n-6 PUFA seen in vivo. We have compared the PUFA profiles of hamster astrocytes cultured in medium supplemented with long-chain PUFA [DHA and/or arachidonic acid (AA)] with those of brain tissue from hamsters fed an n-6/n-3 PUFA-balanced diet or one lacking n-3 PUFA. Astrocytes were obtained from the brain cortex of newborn hamsters and cultured in minimum essential medium + 5% fetal calf serum (FCS) supplemented with DHA and/or AA for 10 days. The astrocytes cultured in medium + FCS had low n-3 PUFA contents, comparable to those of brain tissue from hamsters fed an n-3-deficient diet. We have shown that astrocytes grown in medium supplemented with DHA and/or AA, plus α-tocopherol to prevent lipid peroxidation, incorporated large amounts of these long-chain PUFA, so that the n-6/n-3 PUFA compositions of the phosphatidylethanolamine and phosphatidylcholine, the two main classes of membrane phospholipids, were greatly altered. Astrocytes cultured in medium plus DHA had a more physiological n-3 status, grew better, and retained their astrocyte phenotype. Thus astrocytes in culture are likely to be physiologically relevant only when provided with adequate DHA. This reliable method of altering membrane phospholipid composition promises to be useful for studying the influence of n-6/n-3 imbalance on astrocyte function. © 2003 Wiley-Liss, Inc.