Male rats were treated with one ethanol (2.0 g/kg i.p.) or saline injections once a week for 50 weeks. During this treatment period the rats had in addition access to ethanol (10% in drinking fluid) as a choice against water for 24 h prior to the injection. During the following evaluation period, animals had a continuous choice between ethanol and water and the concentration of the ethanol solution increased every 3rd week from 5 to 10, 15 and 25%, with 10% as a reference tested between the other concentrations. The animals were killed after an abstinence of 4 weeks, whereupon the concentrations of noradrenaline (NA), dopamine (DA), serotonin (5-HT) and 5-hydroxyindole acetic acid (5-HIAA) were determined in the frontal cortex. In the remaining cerebral cortex, activity of monoamine oxidase, reuptake of NA and stimulated inositol phospholipid (PI) breakdown was also determined. Muscarinic binding sites were determined in the striatum. During treatment, saline injected rats had a constant voluntary 24 h ethanol intake. There was a decrease in the corresponding intake in the animals given the ethanol injections. The diminishing of the intake was more marked in rats starting treatment at an age of 19.4 weeks when compared to rats starting at an age of 5.4 weeks. In the evaluation period the ethanol intake was fairly constant for all groups. However, the regressions between intake of the reference concentration when plotted against the different tested concentrations were most marked in the group where ethanol injections started at an early age. In the total material there were significant F-values when concentrations of NA, 5-HIAA, 5-HT/5-HIAA in the cortex and muscarinic binding sites in the striatum were tested. Age could not be excluded as a contributing factor, but for muscarinic binding sites in the striatum, concentrations of DA and 5-HIAA in the cortex, and potassium stimulated PI breakdown in the cortex significant regressions with voluntary ethanol intake as dependent variable could be established. Since these intakes are stable, a causal relation with dependence may be involved.