It has previously been shown that the activity of some areas of the forebrain of birds is dependent on the arousal level of the animal. Other areas do not show this dependency. This paper, on the basis of 2-DG experiments and spine density measurements on Golgi-impregnated tissue, shows that primary telencephalic target areas of the two visual pathways of zebra finch males are not dependent on arousal for activation. In contrast, secondary areas of both visual pathways show arousal-dependent activation. Only the secondary visual areas also show effects of rearing conditions on the spine density: isolation of the birds from day 40 reduces spine density in the hyperstriatum accessorium (HA) of the thalamofugal pathway, and enhances spine density in the lateral neo/hyperstriatum (LNH, tectofugal pathway) significantly from day 80, if compared to aviary-reared birds. A 1-week exposure to a female eliminates the isolation effects in both areas. A second isolation period again reduces the spine density in the HA, but does not enhance it again in the LNH. By comparison with previous studies, we conclude that the spine density in the HA reflects the complexity of the social environment. The irreversible reduction of spine density in the LNH as consequence of the 7-day exposure to a female is interpreted as physiological correlate of an imprinting process, which has previously been shown to occur at the same time. The effects in both areas, the HA and LNH, are dependent on arousal, which may be mediated by brainstem efferents innervating the secondary, but not the primary, visual areas in birds and in mammals.