Neonatal estradiol injections masculinize both the singing behavior and some morphological aspects of the song control nuclei in female zebra finches (Taeniopygia guttata), but there is to date no physiological evidence that estrogens are responsible for the differentiation of song and the song control nuclei in male zebra finches during normal ontogeny. Injections of estrogen into young males demasculinizes their copulatory behavior, so that it is difficult to understand how the same endocrine environment could simultaneously masculinize song and demasculinize copulation. Attempts to block early estrogen action in males have resulted in a paradoxical hypermasculinization of several aspects of the song control nuclei. In the present experiment, we tried to evaluate the role played by estrogens during zebra finch differentiation by blocking their secretion with the non-steroidal aromatase inhibitor, R76713 or racemic vorozole. Zebra finches received at 2-3 days post-hatch one Silastic implant filled with R76713 or left empty as a control. Implants were left in place until about 45 days of age and birds were then gonadectomized. At the age of 105 days, all birds received a Silastic implant filled with T and, starting two weeks later, they were repeatedly tested for singing and copulatory behavior during a period of 3 weeks. They were then perfused and their brains were analyzed by histological methods to measure the volume of the song control nuclei, HVc, RA, area X and MAN. Treatment with R76713 significantly decreased (± 50%) the number of song bouts produced by the birds but did not affect the copulatory behavior of the males. The behavior of females was not changed by the aromatase inhibitor and, in particular, no increase in mounting behavior was seen in the treated females. The presence of a strong sex dimorphism in the size of all 4 song control nuclei was confirmed but no significant effect of R76713 on these measures could be detected. These data are therefore consistent with the idea that estrogens are implicated in the differentiation of singing behavior in the zebra finch but the small amplitude of the behavioral effect observed and the absence of morphological effect suggest that either other factors also play a major role at this level or that higher doses or longer treatments with the aromatase blocker should be tested.