More than 30 years have passed since the publication ( I ) showing that transplacental administration of testosterone to female guinea-pigs caused permanent changes in their ability to exhibit reproductive behaviours in response to ovarian steroids in adulthood. Over the intervening years articles have appeared which have extended this type of observation to numerous other mammalian species, including rat, hamster, gerbil, mouse, dog, ferret, pig, sheep, rhesus monkey, and marmoset (2, 3). In many instances inferences have been made about the probable existence of steroid-induced changes in the structure and function of the nervous system which underlie these changes in behaviour. These so-called 'organizational' effects of perinatal steroid exposure on the developing nervous system are often distinguished from the later 'activational' effects of these same steroids on the expression of sociosexual behaviours in adulthood. The number of different mammalian species used for mechanistic studies on the organizational effects of steroids, such as determining the precise perinatal timing of hormone action, the identity of the steroid(s) responsible for behavioural sexual differentiation, and the neural sites of steroid action, has been relatively small, and generally restricted to rodents. In the late 1970's sex steroids acting during development were first implicated in the differentiation of sexually dimorphic telencephalic structures in passerine song birds which control singing associated with courtship (4, 5). Since this early work, numerous studies have exploited the canary and zebra finch to explore mechanisms of steroid action in controlling the ontogeny of song in the male and the underlying changes in the nervous system. Other studies using the rat have explored in detail the role of androgen and oestrogen in controlling the differentiation in males of patterns of reproductive behaviour (2) as well as aspects of neural sex dimorphisms in the preoptic area/anterior hypothalamus (POA/AH) (6) and in the lumbar spinal cord (7). The concentration of mechanistic (behavioural, biochemical, neuroanatomical) studies of steroidal effects on sexually dimorphic behavioural potential in a few avian and rodent species is understandable, in light of their relative availability, low expense, and rapid time-course of neural development.Despite the substantial progress already made using these lower vertebrate models in promoting our understanding of how sex steroids influence neural and behavioural development, extrapolation of the results obtained in these studies to higher mammals including man cannot occur in the absence of some degree of empirical verification. Obviously, experimental manipulations of steroidal exposure during the development of human subjects would be unethical, thus we are left with experiments of nature (e.g. congenital adreno-genital syndrome,androgen insensitivity syndrome, and Sa-reductase deficient pseudohermaphroditism; details in later sections) with which to ask questions about the role of sex...