The activity of gamma-aminobutyrate aminotransferase (GABA-T) was measured in the brains of rats treated both acutely and sub-chronically with ethanol. Previously, chronic treatment with ethanol for 90 weeks was found to increase the mean brain GABA-T activity by 20-45%. In the present study acute ethanol treatment (4 g/kg, i.p.) did not induce and change in the activity of brain GABA-T with the exception of a small increase in the cerebellum (8%) and, after repeated treatment with ethanol (4 g/kg/day, i.p.) for one and two weeks, no change in the activity of GABA-T was also found in any of the brain regions examined. Subchronic treatment with ethanol for 14 weeks, performed according to two different schedules involving a voluntary intake of ethanol in the drinking water, resulted in approximately a two-fold difference in ethanol intake. A mean increase of 50-85% in the activity of GABA-T was found in all the brain regions of rats with higher ethanol intake in comparison with the group of rats with lower ethanol intake. A bimodal distribution of brain GABA-T activity, however, was found in the ethanol-treated rats, with 60% of the rats having a two-fold increase and the remaining 40% having unchanged activities. The addition of pyridoxal phosphate to the incubation media increased the activity of brain mitochondria from ethanol-treated rats with high brain GABA-T, whereas there was a decrease in the activity in control rats and in ethanol-treated rats in which no increase in brain GABA-T had occurred. These results show firstly, that in a subpopulation of rats, subchronically treated with ethanol for 14 weeks, there was a two-fold increase in brain GABA-T activity, while in another subgroup no change occurred, and, secondly, that this increase in GABA-T activity was a consequence of a change in the response of the apoprotein to the addition of the cofactor pyridoxal phosphate.