Spermatogenesis and male fertility are dependent on FSH, LH, and high levels of testosterone within the testicles. LH does not seem to be involved in spermatogenesis outside of its role in inducing production of testosterone by the Leydig cells in the seminiferous tubules (which make up approximately 80% of the bulk of the testes), whereas this is not the case for FSH, which is importantly involved. In accordance with the fact that the testes are the source of 95% of circulating testosterone in the body, local levels of testosterone inside of the testes are extremely high, ranging from 20- to 200-fold higher than circulating concentrations. Moreover, high levels of testosterone within the testes are required for spermatogenesis, although only a small fraction (5–10%) of normal levels appears to actually be necessary for spermatogenesis.
Unlike with antigonadotropic antiandrogens like CPA and GnRH analogues, it has been reported that bicalutamide monotherapy (at 50 mg/day) has very little or no effect on the ultrastructure of the testes and on spermatogenesis in men even after long-term therapy (>4 years). This may be explained by the extremely high local levels of testosterone in the testes, in that it is likely that systemic bicalutamide therapy is unable to achieve concentrations of the drug within the testes that are able to considerably block androgen signaling in this part of the body. This is particularly so considering that bicalutamide increases circulating testosterone levels, and by extension gonadal testosterone production, by up to two-fold in males, and that only a small fraction of normal intratesticular testosterone levels, and by extension androgen action, appears to be necessary to maintain spermatogenesis. Bicalutamide monotherapy at 50 mg/day causes no or clinically unimportant Leydig cell hyperplasia.
In contrast to bicalutamide and other pure antiandrogens or NSAAs, antigonadotropic antiandrogens suppress gonadotropin secretion, which in turn diminishes testosterone production by the testes as well as the maintenance of the testes by FSH, resulting in atrophy and loss of their function. As such, bicalutamide and other NSAAs may uniquely have the potential to preserve testicular function and spermatogenesis and thus male fertility relative to alternative therapies. In accordance with this notion, a study found that prolonged, high-dose bicalutamide treatment had minimal effects on fertility in male rats. However, another study found that low-dose bicalutamide administration resulted in testicular atrophy and reduced the germ cell count in the testes of male rats by almost 50%, though the rate of successful fertilization and pregnancy following mating was not assessed. Additional studies found that bicalutamide decreased testes weights, altered testes histology, and decreased sperm count in male rats. Yet another study found that bicalutamide has no effect on testes weights or spermatogenesis in male rats.
Treatment of men with exogenous testosterone or other AAS results in suppression of gonadotropin secretion and gonadal testosterone production due to their antigonadotropic effects or activation of the AR in the pituitary gland, resulting in inhibition or abolition of spermatogenesis and fertility:
Treatment of an infertile man with testosterone does [not] improve spermatogenesis, since exogenous administrated testosterone and its metabolite estrogen will suppress both GnRH production by the hypothalamus and luteinizing hormone production by the pituitary gland and subsequently suppress testicular testosterone production. Also, high levels of testosterone are needed inside the testis and this can never be accomplished by oral or parenteral administration of androgens. Suppression of testosterone production by the leydig cells will result in a deficient spermatogenesis, as can be seen in men taking anabolic–androgenic steroids. 
In contrast, pure AR antagonists would, in theory, result in the opposite (although reduced semen volume and sexual dysfunction may occur):
It is theoretically a sound hypothesis that the spermatogenesis can be increased by indirectly stimulating FSH and LH secretions from the pituitary gland. However, for this to fructify, it requires the use of testosterone antagonist to nullify the negative feedback effect of circulating testosterone on the release of FSH and LH, thus augmenting the secretion of testosterone and spermatogenesis. Unfortunately, a testosterone antagonist will be unacceptable to males, as it may reduce secondary sexual functions including erection and ejaculation that is vital for the successful fertilization. 
However, while bicalutamide does not appear to adversely influence testicular spermatogenesis, and healthy sperm can be produced within the testes during bicalutamide monotherapy, AR antagonists may be able to interfere with male fertility via interference with androgen signaling beyond the testes. The maturation as well as transport of sperm occurs not only in the testes but also outside of the testes in the epididymides and vas deferens, and these processes in these tissues are dependent on AR signaling similarly to testicular spermatogenesis. However, whereas androgen levels are extremely high in the testes, this is not true in the epididymides and vas deferens. As androgen levels are relatively low in these tissues, at least compared to the testes, bicalutamide may be able to block AR signaling in these parts of the body to an extent that is sufficient to interfere with male fertility. Indeed, the AAS mesterolone has been used to improve sperm quality and fertility in men because, apparently unlike other AAS, it shows minimal antigonadotropic effects at typical clinical dosages but activates the AR and thereby supports sperm maturation in the epididymides. However, this use of mesterolone is controversial and its efficacy for such purposes is not fully certain.
Although bicalutamide alone would appear to have minimal detrimental effect on testicular spermatogenesis and hence on certain aspects of male fertility, other hormonal agents that bicalutamide may be combined with, including GnRH analogues and particularly estrogens (as in transgender hormone therapy), can have a considerable detrimental effect on fertility. This is largely a consequence of their antigonadotropic activity. Antigonadotropic agents like high-dose CPA, high-dose androgens (e.g., testosterone esters), and GnRH antagonists (though notably not GnRH agonists in the case of fertility) produce hypogonadism and high rates of severe or complete infertility (e.g., severe oligospermia or complete azoospermia) in men. However, these effects are fully and often rapidly reversible with their discontinuation, even after prolonged treatment. In contrast, while estrogens at sufficiently high dosages similarly are able to produce hypogonadism and to abolish or severely impair spermatogenesis, this is not necessarily reversible in the case of estrogens and can be long-lasting after prolonged exposure. The difference is attributed to an apparently unique, direct cytotoxic and adverse effect of high concentrations of estrogens on the Leydig cells of the testes.