Testosterone is a significant biological chemical contributing to mammalian development. Early levels of exposure to it determines permanent behaviour. Higher levels of testosterone is predominantly present in males, generating some of the characteristics associated with gender stereotypes such as choice of toys as children and sexual orientation. The variation in the levels of testosterone also had an impact, such as in females with high levels of prenatal exposure to testosterone leading to congenital adrenal hyperplasia (“CAH”). Women with high levels of prenatal exposure to testosterone who develop CAH have shown more typically male behaviours, such as a preference for stereotypically male toys, more interest in rough play. This behaviour is more present than in the female born without CAH, but not as present as it is in individuals born male. This supports the proposition that gender is biological and not social as the levels of exposure to testosterone prenatally determines male or female behaviours. 
Testosterone cannot be used to argue whether someone is more male or female. It is known to exist in high quantities in males, and significant research has been done to associate high levels with male characteristics that associate with gender roles. This finding has been used to explain and even defend stereotypically male behaviour. There is little, if any, research, however, in the dynamics of testosterone levels and female aggression. We therefore do not know enough about testosterone to conclude that it causes and drives male aggression. Aggression is simply classified as male behaviour and testosterone a male hormone. Testosterone therefore cannot be used to categorise individuals as male or female.  Women who are athletic are associated with having higher levels of testosterone and therefore being more genetically male. Testosterone levels can influence muscles, but this varies from person to person. Some people make better athletes with lower levels.
Rejecting the premises