Some eunuchs can still get erections, though the ability is significantly diminished compared to intact men. In a study of sixteen castrated men exposed to visual erotic stimuli, a quarter of them achieved a functional erection, and the men who did had measurably higher residual testosterone levels than those who did not. The answer depends heavily on factors like age at castration, how much time has passed, whether the body has other sources of androgens, and what type of stimulation is involved.
Why Castration Impairs Erections but Does Not Always Eliminate Them
The testes produce the vast majority of a man’s testosterone, and testosterone is deeply involved in maintaining the tissues and nerve signaling that make erections possible. Without it, the smooth muscle inside the erectile chambers of the penis gradually thins out and gets replaced by connective tissue and fat-containing cells. This structural remodeling makes the tissue stiffer and less capable of trapping blood, which is the physical basis of an erection.
Animal research has consistently shown that castration reduces the pressure achieved inside the erectile tissue when the relevant nerve is stimulated. In castrated dogs, the threshold of nerve stimulation needed to produce any erection was much higher than in intact animals, and the resulting erections were weaker: the pressure inside the erectile chambers reached only about 57% of peak blood pressure, compared to 80% in intact dogs. Similar results appear across species, with castrated rabbits and rats all showing reduced but not zero erectile responses to nerve stimulation.
The reason erections weaken rather than vanish entirely is that erection is not a single testosterone-controlled switch. It involves local nerve signals, blood vessel dilation driven by nitric oxide, and direct physical stimulation pathways, some of which continue to function even when testosterone drops to very low levels. Testosterone is best understood as a maintenance hormone for the erectile system: it keeps the parts in good working order, but those parts do not all shut down instantly when the hormone disappears.
Nocturnal and Spontaneous Erections Are Hit Hardest
Not all erections depend on testosterone to the same degree. Research comparing hypogonadal men (men with very low testosterone, functionally similar to castrated men) with healthy controls found that erections during sleep were substantially reduced in the low-testosterone group, both in size and in rigidity. These nocturnal erections are strongly androgen-dependent and tend to be the first casualty when testosterone drops.
Erections triggered by visual erotic stimuli, on the other hand, told a different story. In the same studies, the circumference increase in response to erotic imagery was similar between hypogonadal men and controls. Rigidity was also comparable, though testosterone replacement did improve both the duration and maximum rigidity of these stimulus-driven erections. The researchers concluded that erections triggered by external erotic cues primarily involve a system that does not depend on androgens, but that system may still benefit from testosterone’s support.
This distinction matters for understanding what a castrated man actually experiences. Spontaneous morning erections and the erections that happen during REM sleep are likely to fade or disappear. But erections in response to direct physical stimulation or arousing situations may persist to some degree, especially in the months or years immediately following castration before the tissue changes become severe.
Evidence from Castrated Men
Direct studies of castrated humans are rare for obvious ethical reasons, but the data that exists is revealing. In a study of sixteen men who had been surgically castrated, researchers measured erectile responses to visual erotic stimulation. Four of the sixteen, or 25%, achieved what was classified as a functional erection. The men who achieved erections had mean free testosterone levels of about 1.1 pg/mL, while those who did not averaged about 0.6 pg/mL. Both values are extremely low compared to normal male ranges, but that small difference was statistically significant and apparently mattered for function.
Where was that residual testosterone coming from? The adrenal glands, which sit atop the kidneys, produce small amounts of androgens including DHEA and androstenedione, which the body can convert into testosterone. This adrenal contribution is normally trivial compared to what the testes produce, but after castration it becomes the only source. In rat experiments, removing both the adrenal glands and the testes produced a much more severe loss of erectile function than removing the testes alone: the erectile pressure ratio dropped to 0.16 in doubly-deprived animals, compared to 0.40 with castration alone and 0.75 in intact controls. When researchers gave the doubly-deprived rats adrenal hormones back, erectile response recovered to near-normal levels.
This tells us that adrenal androgens serve as a partial safety net. The men in the castration study who managed erections may have been the ones whose adrenal glands were producing enough androgens to keep the erectile machinery minimally functional.
Age at Castration Makes a Big Difference
A man castrated before puberty never develops the full adult penis and erectile tissue architecture that testosterone builds during adolescence. The smooth muscle, blood vessel network, nerve fiber density, and structural proteins that make erection mechanically possible all depend on pubertal testosterone exposure for their development. Androgens regulate the differentiation of precursor cells into the trabecular smooth muscle that fills the erectile chambers, maintain the nerve fibers that trigger dilation, and support the structural integrity of the surrounding tissue.
A man castrated after puberty, by contrast, starts with fully developed erectile hardware. The decline is gradual: tissue remodeling takes months to years, and the neural pathways that were established during development do not immediately disappear. Historical accounts support this timeline. A cross-cultural survey examining eunuchs across Byzantine, Roman, Ottoman, Chinese, and Italian castrati societies found that rather than being uniformly asexual and celibate, eunuchs were often sexually active. Many of those eunuchs were castrated as boys, but the accounts of sexual activity are more consistent among those castrated as adults or adolescents, when some erectile development had already occurred.
What Happens to Desire vs. Physical Capability
Libido and erectile function, while related, are not the same thing. Testosterone has a strong influence on sexual desire, and most castrated men report a dramatic drop in spontaneous sexual thoughts and interest. But the mechanical apparatus for erection is partly independent of the desire system. A man can lose most of his libido while retaining some physical erectile capability, or vice versa.
This dissociation is well documented in prostate cancer treatment, where androgen deprivation therapy (chemical castration using drugs that suppress testosterone to castrate levels) is commonly used. About 20% of men on androgen deprivation therapy retain some degree of both libido and erectile function. The majority experience substantial or total loss of both, but the minority who retain function demonstrate that the relationship between testosterone, desire, and erections is not a simple on-off switch.
The brain also contributes pathways that do not depend on testosterone. Dopamine signaling, particularly through certain receptor types in the hypothalamus and limbic brain regions, plays a major role in sexual arousal and the neural commands that initiate erections. These dopaminergic circuits are involved in the preparatory phase of sexual behavior, including arousal and motivation, and in the hypothalamic pathways that connect to the spinal cord centers controlling penile erection. Because these circuits are not solely testosterone-dependent, they represent another route through which some erectile response can persist after castration.
Do Erectile Medications Still Work?
Drugs like sildenafil (Viagra), tadalafil (Cialis), and vardenafil work by blocking an enzyme called PDE5, which breaks down the chemical signal that keeps erectile tissue relaxed and filled with blood. The problem for castrated men is that testosterone regulates how much PDE5 the erectile tissue produces. When testosterone is very low, PDE5 expression changes, and the drugs may have less to work with. A systematic review found that in men with low or low-normal testosterone who did not respond to PDE5 inhibitor treatment alone, adding testosterone improved sexual function.
So the short answer is that erectile medications are less likely to work in a castrated man than in a man with normal testosterone levels, but they are not guaranteed to fail either. Normal and high testosterone levels are associated with increased nitric oxide synthase and the downstream signaling molecules that PDE5 inhibitors amplify. Without testosterone, those upstream signals are weakened, which reduces the drugs’ effectiveness. In men whose testosterone is merely low rather than essentially zero, the combination of testosterone replacement and a PDE5 inhibitor can sometimes rescue erectile function that neither treatment alone could achieve.
For a fully castrated man with no testosterone replacement, the likelihood of a robust response to Viagra alone is low. But some response is possible, particularly if adrenal androgen production is still providing a baseline of hormonal support.
Can Testosterone Replacement Reverse the Damage?
In animal studies, giving testosterone back to castrated animals consistently improves erectile function, sometimes dramatically. In castrated rabbits, testosterone treatment prevented the effects of castration and restored erectile pressure to values similar to intact animals, while estradiol did not. In rats, testosterone therapy produced a substantial increase in erectile response compared to untreated castrated animals.
Timing matters, though. Rat studies looking at delayed testosterone replacement found that when treatment was started late after castration, four weeks of testosterone was not enough to restore function. Eight weeks did produce significant improvement, but the erectile response still did not fully return to pre-castration levels. The tissue that was lost to remodeling, the smooth muscle replaced by connective tissue and fat, does not simply snap back. Recovery is possible but incomplete, and the longer the gap between castration and treatment, the harder recovery becomes.
The nitric oxide signaling system, which is the chemical pathway that triggers blood vessel dilation in the penis, also shows testosterone dependence that is at least partly reversible. In a rat study, the density of the nerve fibers that produce nitric oxide decreased proportionally to the time after castration. When testosterone was restored, both the nerve fiber density and the erectile response returned to near-normal values. This reversibility is encouraging, but “near-normal” in a controlled animal experiment may still translate to noticeably impaired function in a human being going about daily life.
Modern Clinical Parallels
The most informative modern parallel comes from transgender women who undergo bilateral orchiectomy (surgical removal of both testes) as part of gender-affirming care, typically while also taking estrogen and anti-androgen medications. A study examining this population found that orchiectomy combined with hormone therapy significantly reduced spontaneous erections compared to hormone therapy alone. However, and this is the key finding, orchiectomy did not eliminate erections associated with sexual arousal and activity. Over half of participants in both the hormone-therapy-only group and the hormone-therapy-plus-orchiectomy group reported experiencing erections with sexual arousal after treatment.
This modern clinical data aligns with the older research on castrated men and hypogonadal patients: the erectile system is weakened but not destroyed by removing the testes. The erections that survive tend to be the ones driven by active sexual stimulation rather than the spontaneous or nocturnal ones that depend more heavily on circulating androgens.
There is also an important behavioral variable at play. In the transgender women studies, those who continued to use their penis for sexual activity maintained higher erectile function scores than those who did not. Use it or lose it appears to apply: the tissue remodeling that castration triggers may progress faster when the erectile tissue is not being regularly engaged. Regular erections bring blood flow and mechanical stretching to the tissue, which could slow the loss of smooth muscle and the accumulation of connective tissue and fat.
How the Adrenal Glands Act as a Backup System
The adrenal contribution deserves more attention because it explains much of the variability between castrated individuals. Everyone’s adrenal glands produce androgens, but the amount varies from person to person. After castration, the adrenal output becomes the entire androgen supply. In the rat experiment that tested this directly, castration alone reduced the erectile pressure ratio from 0.75 to 0.40, meaning about half the normal response remained. Adding adrenalectomy dropped it further to 0.16. Replacing the adrenal hormones (with hydrocortisone and aldosterone) brought the ratio back up to 0.63, remarkably close to normal.
This suggests that in humans, individual variation in adrenal androgen production could explain why some castrated men retain reasonable erectile function while others lose it almost entirely. Factors like stress, overall health, adrenal gland size, and genetic variation in androgen-producing enzymes all influence how much testosterone and its precursors the adrenal glands contribute. A castrated man with robust adrenal output might retain enough androgenic support to maintain partial erections for years, while one with lower adrenal production might lose function quickly.
The Historical Record and What It Tells Us
Historical accounts of eunuchs provide a rough, uncontrolled but culturally rich dataset. Across widely separated civilizations spanning centuries, eunuchs were documented as sexually active individuals, not uniformly celibate. The cross-cultural survey examining eunuch history across six major historical periods and cultures found that many eunuchs were recognized not only for political roles and administrative skill but also for maintaining sexual relationships. Some served in harems specifically because they were believed to be safe around women, yet historical sources frequently note that this assumption was not always accurate.
The historical evidence is obviously imprecise. It is impossible to know what fraction of historical eunuchs retained erectile function, at what age they were castrated, or how their subjective experience compared to intact men. But the consistency of reports across unrelated cultures, from Byzantium to China to the Ottoman Empire, makes it clear that castration did not reliably eliminate all sexual capability. The historical record tracks with the modern clinical data: some eunuchs retained function, many did not, and the outcome varied by individual.
The Nitric Oxide Connection
At the molecular level, one of the most important things testosterone does for erections is maintain the enzyme system that produces nitric oxide in penile tissue. Nitric oxide is the key chemical signal that tells the smooth muscle in the erectile chambers to relax, allowing blood to rush in. The enzyme that produces it, nitric oxide synthase, exists in both the nerve fibers and the blood vessel lining within the penis. Testosterone supports both forms.
A rat study that tracked erectile response and nitric oxide synthase activity over time after castration found that nerve fiber density producing nitric oxide decreased proportionally with time since castration, and erectile response declined in parallel. When testosterone was restored, both the nerve fibers and the erections recovered. In human erectile tissue studied in the lab, normal and high testosterone levels were associated with increased expression of both the neuronal and endothelial forms of nitric oxide synthase, along with higher levels of the downstream signaling molecule that produces smooth muscle relaxation.
One complicating finding from a mouse castration study was that nitric oxide synthase gene expression did not always change significantly after castration, which suggests the picture may be more complex than a simple “testosterone turns on NOS, no testosterone turns it off” model. The relationship likely involves multiple layers of regulation including protein-level changes, oxidative stress, and nerve fiber survival rather than gene expression alone.
What this means practically is that the erectile tissue of a castrated man is not fundamentally broken at the circuit level. The wiring is still there. The chemical signaling pathway still exists. But without testosterone maintaining the system, the signals get weaker, the tissue gradually remodels into something less responsive, and the threshold for achieving an erection climbs higher and higher. For some men, enough residual androgen and enough direct stimulation can still clear that higher threshold. For many, it cannot.