European starling's (Sturnus vulgaris L.) were used to measure differences in the glucose metabolism in the auditory forebrain between birds performing an auditory discrimination task and birds habituated to the same acoustic stimuli. One group(n = 5) of individuals was trained in an operant GO/NOGO-procedure to report 1-kHz tone signals in a background of 4-kHz stimuli. The other group(n = 5) was habituated to the experimental set-up and to the same sequence of tones presented to the trained birds. [ 14 C]2-deoxyglucose (2DG) uptake was determined in the caudal auditory telencephalon and the nucleus ovoidalis of well trained and habituated birds by autoradiography of brain sections. The tissue areas having grey values above predefined threshold values of labelling were determined in every brain section of each bird and then combined to volumes of labelled tissue. No significant differences of the 2DG uptake in the nucleus ovoidalis were found between the two experimental groups. In the caudal auditory telencephalon, however, significant differences in 2DG-labelling were found. In the trained birds, the labelling in the caudal auditory telencephalon was confined to smaller brain regions than in the habituated birds. These results suggest a differential processing of sounds in the trained and habituated birds which is discussed in the context of sharpening of the frequency representation by GABAergic inhibition and processes of attention.