The molecular mechanism of volatile anesthetic action remains unknown. Attempts to elucidate this mechanism have been complicated by the absence of models in which changes in neuronal cellular properties can be correlated with changes in whole animal anesthetic effect. In this study we describe a model where dietinduced alterations in rat brain fatty acid composition are correlated with alterations in volatile anesthetic potency. Rats maintained on a fat-free diet showed significant depletion of arachidonic acid (20:4&6; 5,8,11,14-eicosatetraenoic acid) and docosahexaenoic acid (22:6w3; 4,7,10,13,16,19,-docosahexaenoic acid) in brain, and a corresponding increase in Mead acid (20: 3w9; 5,8,11-eicosatrienoic acid). These fat-deprived rats were significantly more sensitive to all volatile anesthetics tested than were age-controlled rats on a normal diet. Parenteral supplementation of the fat-deprived animals with linolenic acid (18: 303, 9,12,15-octadecatrienoic acid) completely reconstituted the docosahexaenoic acid content of brain without affecting anesthetic sensitivity. In contrast, supplementation of the fat-deprived rats with linoleic acid (18:2w6; 9,12-octadecadienoic acid) caused a dramatic decrease in anesthetic sensitivity, but only a small change in whole brain arachidonate content. Further analysis revealed that linoleate supplementation of fat-deprived animals resulted in a preferential noraiztion of the arachidonate content of brain phosphatidylinositol as compared with other brain phosphoglycerides. These results demonstrate for the first time a correlation between changes in membrane composition and anesthetic effect, and indicate that the precise fatty acid composition (perhaps in specific phospholipids) of brain is important in the mechanism of volatile anesthetic action.