It is currently accepted that most sex differences in brain and behavi
or do nor result from direct genomic actions, but develop following ea
rly exposure to a sexually differentiated endocrine milieu. In Japanes
e quail (Coturnix japonica), in contrast to rodents, the male reproduc
tive phenotype appears to develop in the absence of endocrine influenc
e, and estradiol secreted by the ovary of the female embryo is respons
ible for the physiologic demasculinization of females. In zebra finche
s (Taeniopygia guttata), estrogens administered early in life demascul
inize copulatory behavior in males, but masculinize the vocal control
regions in the brain and singing behavior of females. It is difficult
to understand how these behaviors differentiate given that normal untr
eated males sing and copulate in a male-typical manner, whereas female
s never show these behaviors. All attempts to resolve this paradox wit
h experiments based on the rodent model of sexual differentiation have
been unsuccessful. We propose that copulatory behavior in zebra finch
es is differentiated in a manner similar to what has been described in
quail, but that novel approaches need to be considered to understand
the differentiation of the telencephalic song control system. In parti
cular the possible involvement of afferent input that may differentiat
e in a steroid-dependent or -independent manner should be thoroughly t
ested.