You tell whether hormone therapy is working by what your symptoms do over about three months. In the hormone-therapy service reviews we publish here, readers ask about retesting policy constantly. Those services differ far more on how often they draw blood than on what they prescribe.
Hormone replacement therapy (HRT), also called menopausal hormone therapy, is judged on how you feel and on whether side effects show up. That sounds imprecise next to a laboratory value. It is the standard the guidelines set, because no serum number has been shown to correspond to a good outcome.
The Verdict
Symptom Relief Is the Measure the Guidelines Use
Relief of menopausal symptoms without adverse effects signals an adequate response to hormone therapy. StatPearls records that routine testing of follicle-stimulating hormone, estradiol or progesterone is not recommended for monitoring or directing treatment, and attributes that position to the American College of Obstetricians and Gynecologists (ACOG). The Menopause Society describes the practical consequence on its patient page. A period of trial and error is usually needed to arrive at the best dose.
ACOG takes the same line before treatment starts. Hormone levels move too much during the menopause transition for one measurement to mean much.
This matters for what you should write down. A symptom count you keep yourself carries more weight in a follow-up appointment than a laboratory printout does. Count hot flashes over three days, note how many times you wake, and rate the two symptoms that bother you most. Do that before you start, then again at three months. The UK's National Institute for Health and Care Excellence (NICE) recommends a review at that point for how well the treatment works and how well it is tolerated. Reviews run annually after that.
How Long Each Symptom Takes to Respond
Different symptoms respond on different clocks, and the gap between the fastest and the slowest is months rather than weeks. Vasomotor symptoms, meaning hot flashes and night sweats, move first. Mood and joint pain move last, and the trial record behind that one is modest and inconsistent rather than absent.
| Symptom | Change usually starts | When to judge it | Common mistake |
|---|---|---|---|
| Hot flashes and night sweats | 2 to 4 weeks | About 3 months | The frequency usually drops before the intensity does |
| Broken sleep | 2 to 6 weeks | About 3 months | Often improves as night sweats settle rather than on its own |
| Vaginal dryness and painful sex | 4 to 8 weeks | About 3 months | Local vaginal estrogen works faster here than a patch or pill |
| Low mood and irritability | 4 to 12 weeks | About 3 months | The slowest of the common symptoms to shift, and the least specific |
| Joint and muscle aches | 8 to 12 weeks | About 6 months | Improvement is partial for most women rather than complete |
| Bone density | Nothing you can feel | 12 to 24 months on a scan | The benefit is real and invisible, which is why it gets abandoned |
Most women see hot flash frequency fall noticeably by week four, with the improvement continuing well past that point. Do not judge the result there. A treatment that will work often produces only a partial response at four weeks. Stopping there costs people a dose that would have suited them.
The bone row is the one worth sitting with. Hormone therapy reduces fracture risk, and you will never feel it happening. If bone protection is part of why you started, a repeat scan at 12 to 24 months is the only thing that will show you whether that part is working.
Why Blood Levels Do Not Tell You If Hormone Therapy Is Working
Serum estradiol correlates poorly with symptom relief, which is the practical reason no target range exists. Women reach comfort at widely different levels, and the same reading in two people can sit alongside completely different symptom burdens. There is also no single premenopausal figure to aim at, because estradiol swings from roughly 20 to 350 pg/mL across a normal cycle on mass-spectrometry reference intervals. Dosing to reach a number tends to produce side effects rather than better control.
Follicle-stimulating hormone has the same problem in a different form. It falls somewhat once estrogen therapy starts, but the size of the fall does not track how well you feel. During perimenopause it is close to useless on its own, because it can swing across the whole range within a single cycle.
There is a measurement problem underneath the interpretation problem. Routine immunoassays were built for the higher concentrations of reproductive-age women, and they lose accuracy at postmenopausal levels, so two laboratories can return meaningfully different numbers from one draw. The route you take matters here too, because oral estrogen reads falsely low on a widely used immunoassay. Weigh that before a single reading changes a dose that is otherwise controlling your symptoms.
When a Level Is Worth Checking
Four situations make a blood level genuinely useful, and none of them is routine follow-up on a patch that is working.
| Situation | What a level adds | What it still does not tell you |
|---|---|---|
| Pellets or implants | The dose cannot be adjusted once it is in, and levels can climb well above the premenopausal range | Whether your symptoms are about to improve |
| No change at all on a standard dose | Poor absorption is one explanation a level can rule in or out | What your dose ought to be |
| Symptoms and side effects at the same time | Helps separate too little estrogen from a cause that is not hormonal | A target number to aim for |
| Premature ovarian insufficiency | Treatment here replaces what an ovary would have made until the usual age of menopause, so a level has a reference point | How you will feel day to day |
Pellets are the strongest case. Once a pellet is inserted the dose cannot be dialled back. Levels frequently run above anything the other routes deliver, so a blood test is the only way to see where you have landed. Our page on HRT versus compounded bioidentical hormones covers why compounded pellet regimens sit outside the approved products.
Premature ovarian insufficiency is the other genuine exception, and it is a different clinical situation rather than an early version of the same one. Treatment there replaces hormones a woman would still have been producing, so a level has a reference point.
What Biological Age Tests Can and Cannot Show
No biological age test is validated to tell you whether your hormone therapy dose is right. Glycan tests read the sugar chains attached to IgG antibodies, and falling estrogen shifts that pattern toward the more inflammatory form. Epigenetic clocks read DNA methylation instead. Both sets of markers also move with weight, illness, sleep, smoking and infection, so a change in the number cannot be pinned on the hormone therapy.
A marker that moves during treatment is not the same as a marker that can steer it. To steer a dose, a test needs to show that moving the number leads to a better outcome. That evidence does not exist for these tests in menopause care. Marketing that presents a biological age result as a way to titrate estrogen is describing a use the research has not established, and no menopause guideline sets a dose from one.
A baseline biological age test before you start is not worth paying for. It looks reasonable, because a before-and-after pair is how you would test anything. A single follow-up reading cannot separate the effect of estrogen from a winter cold, ten pounds of weight, or three bad months of sleep.
These tests can still be interesting as a general health measure over years. Our explainer on reading a glycan age result and our page on high-sensitivity C-reactive protein (hs-CRP) compared with glycan markers set out what those numbers do support. Treat them as separate from whether your hormone therapy dose is right.
Signs Hormone Therapy Is Not Working
Three months at a standard dose with no change in your main symptom is the clearest signal that something needs to change. A few other patterns are worth raising sooner.
- Symptoms return after a good early response. With patches and gels this often points to absorption rather than dose, and application site, skin condition and heat all affect it.
- New bleeding after the first six months. Any bleeding after that point needs assessment rather than a dose adjustment, and it is the one item on this list that should not wait for a routine review.
- Side effects arriving with no symptom benefit. Breast tenderness, nausea, headaches or swelling alongside unchanged hot flashes usually means the route or the progestogen needs changing rather than the estrogen dose rising.
- Only some symptoms improve. A partial response is real information. It suggests estrogen was part of the problem and something else accounts for the rest.
That last pattern is the one most often mistaken for treatment failure. Thyroid disease, iron deficiency, sleep apnoea and depression all cause fatigue and poor concentration in the same age group, and none of them improves because an estrogen dose went up. Our page on HRT side effects and risks covers which new symptoms warrant a call rather than a wait.
Unscheduled Bleeding in the First Months
Unscheduled vaginal bleeding in the first 3 months of hormone replacement therapy is a common side effect that gets reported to your clinician rather than read as a treatment failure. Symptom relief remains the measure of whether systemic therapy is working.
NICE guideline NG23 states that unscheduled bleeding in the first 3 months should be reported at your 3-month review appointment. If you do not have that review booked, book an appointment with your clinic.
Bleeding that starts after the first 3 months should be reported promptly rather than waiting for a routine review. That 3-month line replaces the six-month threshold given in the list above. In April 2026, NICE amended recommendation 1.8.4 in NG23 and added recommendation 1.8.5. Both align NG23 with the advice on unscheduled vaginal bleeding while taking systemic HRT in NICE's guideline on suspected cancer.
Timing changes what happens next, while bleeding volume does not alter the reporting rule. NG23 notes the evidence is limited for people who get unscheduled bleeding on sequential or continuous HRT.
NICE points to the British Menopause Society, which publishes its own guidance on managing unscheduled bleeding on HRT. Do not change your dose or stop taking your hormones before speaking with your clinician.
Who This Does Not Apply To
This guidance covers hormone therapy taken for menopausal symptoms, and three groups sit outside it.
Gender-affirming hormone therapy is monitored against target ranges, and blood levels are part of standard care rather than an exception to it. Anyone in that situation should follow the monitoring their prescribing service sets rather than the symptom-led approach described here.
Testosterone therapy in women is also monitored on levels, because the risk being watched for is going above the female physiological range. Our page on testosterone therapy for women covers what that monitoring looks like and what the evidence supports it for.
Women with premature ovarian insufficiency are treated to replace what the ovaries would otherwise be producing, usually at higher doses and until the average age of menopause. A level then has something meaningful to be compared against. If you started hormone therapy before about age 40, ask your prescriber which framework you are being managed under. The answer changes what a normal review looks like.
What Would Change This Answer
A trial showing that titrating to a serum estradiol target produces better symptom control or fewer fractures than symptom-led dosing would change it. So would a validated marker of estrogen effect in tissue, which is a different thing from a blood concentration and does not currently exist for routine use. Either would give follow-up something to aim at beyond how you feel.
Until one of those arrives, the practical route is unchanged. Write down a three-day symptom count now, and book the review for three months rather than four weeks. Set the two counts side by side at that appointment, and they will tell you whether your hormone therapy is working.
Frequently Asked Questions
How long does it take to know if hormone therapy is working?
Give it three months before you judge it, which is also when NICE guideline NG23 recommends reviewing menopause treatment for efficacy and tolerability. Hot flashes and night sweats start easing for most women within 2 to 4 weeks, and the full effect builds over 8 to 12 weeks. Sleep and mood lag behind that. Judge at four weeks and you will often change a dose that did not need changing.
Should I get my estradiol level checked to see if my dose is right?
Not for standard patches, gels or tablets. StatPearls records that routine testing of follicle-stimulating hormone, estradiol or progesterone levels is not recommended for monitoring or directing treatment. Symptom relief without side effects signals an adequate response instead. There is no serum number that corresponds to "optimised". A level becomes useful in four specific situations, and those are exceptions rather than routine.
Do FSH levels show whether hormone therapy is working?
No. Follicle-stimulating hormone (FSH) rises as ovarian function declines, and estrogen therapy pushes it partway back down, but the amount it moves does not track how well symptoms are controlled. Two women with the same FSH can have completely different symptom burdens. FSH is also unreliable during perimenopause because it swings week to week, which our page on the perimenopause biomarker panel covers in more detail.
Can a biological age or glycan test show whether hormone therapy is working?
Not in a way that should guide your dose. Glycan tests read the sugar chains on IgG antibodies. Epigenetic clocks read DNA methylation. Both sets of markers shift with weight, illness, sleep and infection as well as with hormones. A marker that moves during treatment is not the same as one validated to titrate it, and no menopause guideline uses either to set a dose. Our guide to reading a glycan age result explains what the number does and does not support.
What if my hot flashes stopped but I still feel awful?
That is common, and it usually means something other than estrogen deficiency is driving the rest. Thyroid disease, iron deficiency, sleep apnoea and depression all produce fatigue and brain fog that look like menopause symptoms and do not respond to hormone therapy. A partial response is useful information rather than a failure, because it tells you estrogen was part of the picture. The next step is testing for the causes it did not fix. Raising the dose does not address them.
What if nothing has changed after three months?
Three months of no change at a standard dose is a reason to go back to the prescriber, and there are three usual explanations. The dose or route may not be delivering enough, which absorption can explain with patches and gels. The symptoms may not be hormonal. Or the preparation may not suit you, and a different route often works where a dose increase does not.
Does the route change how quickly it works?
Route changes the speed a little and the risk profile more. Gels and sprays reach a steady level within a few days, patches within a day or two of the second change, and oral tablets within about a week. Those differences are small next to the 8 to 12 weeks the symptom response takes. Local vaginal estrogen is the exception worth knowing, because it treats dryness and painful sex faster than systemic treatment and does very little for hot flashes.
Is a high estradiol level a sign the dose is too high?
Not by itself. Plenty of women feel well at levels a laboratory flags, and plenty feel unwell inside the reference range. Side effects are the signal that matters: breast tenderness, nausea, headaches, leg swelling or new bleeding. Pellets are the case where a number genuinely helps, because the dose cannot be reduced after insertion and levels can run several times higher than any other route delivers.