Pelvic floor muscle training is the treatment clinical guidelines recommend first for stress urinary incontinence. Here is how it is done, what the research does and does not show, and when it may not be the right approach for you.
Health information online often presents confident claims that the underlying research does not support. Throughout this guide, key claims carry a label showing how strong the evidence behind them is:
Well established Guideline-level or systematic review evidence.
Context-dependent Real evidence, but specific to a particular population or condition type.
Limited evidence Small studies or inconsistent findings. Treat with caution.
Which kind of incontinence this applies to
This matters more than most articles acknowledge, because the evidence differs substantially between types.
- Stress urinary incontinence (SUI) — leakage when you cough, sneeze, laugh, lift or exercise, without a preceding urge. This is the type with the strongest evidence for pelvic floor training, and most of the figures in this article relate to it.
- Urgency urinary incontinence (UUI) — a sudden, powerful need to pass urine, sometimes with leakage before reaching a toilet. Pelvic floor training is used here too, but the evidence is weaker and bladder training is usually part of the approach.
- Mixed urinary incontinence (MUI) — features of both.
If you do not know which you have, that is itself a reason to speak to a clinician. The right approach differs, and guidelines treat these as distinct conditions rather than one problem.
What the pelvic floor does
The pelvic floor is a group of muscles running from the pubic bone at the front to the base of the spine at the back, and between the sitting bones.
It contributes to three things:
- Support for the bladder, uterus and bowel
- Closure of the urethra and rectum
- Rapid response to sudden rises in abdominal pressure
That third function is the relevant one for stress incontinence. When you cough or sneeze, pressure inside the abdomen rises sharply. NICE describes pelvic floor training as working by increasing urethral closure pressure and stabilising the urethra, preventing downward movement during activity.1
Training therefore has two components — building strength and endurance, and improving the timing of the response.
What the evidence actually shows
Well established Pelvic floor muscle training is recommended as first-line treatment for stress or mixed urinary incontinence in NICE guidance.1
The main evidence base is a 2018 Cochrane systematic review covering 31 randomised trials and 1,817 women across 14 countries. Within that review, three trials of 242 women with stress incontinence found that 74% of those doing pelvic floor training reported cure or improvement, compared with 11% of controls.2
Three caveats belong with that number, and are usually left out:
- It relates to stress incontinence specifically, not all types
- Cochrane rated it moderate-quality evidence, not high
- Most included trials followed women for under 12 months, so this says little about long-term outcomes
Pelvic floor training is recommended first because it is low-risk, low-cost and non-invasive — not because it outperforms every alternative. A systematic review of cure rates found surgical procedures for stress incontinence achieved a median cure rate of 82.3%, higher than that reported for pelvic floor training.3
That is not an argument for surgery. It is an argument for understanding what you are choosing and why, and for discussing options with a clinician rather than assuming training is the only route.
Finding the right muscles
Context-dependent You may have read that most women do this wrong. The evidence is more mixed than that suggests.
In a cross-sectional study of 779 women attending community primary care, correct contraction on the first attempt was recorded in 85.5% of women with prolapse, 83.4% with stress incontinence, 68.6% with both, and 85.8% with neither.4 A much smaller and older study of 47 women in a urodynamics clinic found only 49% achieved what researchers classed as an ideal contraction after brief verbal instruction.5
Estimates across the literature therefore range from roughly 15% to 50% contracting incorrectly, depending heavily on the population studied and how contraction was assessed.
The more useful finding: in that 779-woman study, of 120 women who initially contracted incorrectly, 94 (78%) learned to do it correctly after brief instruction.4 Getting it wrong at first is common and usually correctable — but correcting it generally required someone assessing them.
Locating the contraction
Imagine you are trying to stop yourself passing wind and stop the flow of urine at the same time. The movement is an inward and upward lift, not a downward push.
Checking with a mirror
Sitting or lying privately with a hand mirror, a correct contraction generally produces a slight inward drawing of the perineum. Visible bulging or downward movement suggests you are bearing down instead — the opposite of the intended action.
Checking by touch
With a clean finger inserted into the vagina, a correct contraction is usually felt as a gentle squeeze and slight upward movement.
A note on the urine-stop test
Some clinicians use stopping the flow of urine mid-stream as a one-off way to identify the muscles. Others advise against it entirely, because repeatedly interrupting the flow may interfere with normal bladder emptying, and no clear evidence establishes that even a single use is necessary or advisable.
If you use it at all, do not use it as an exercise. If you are unsure whether you have found the right muscles, assessment by a physiotherapist is the reliable method — self-assessment is genuinely difficult.
Common errors
Context-dependent In a trial of 113 pregnant women, the documented causes of incorrect or insufficient contraction were co-contraction of other muscle groups — gluteal muscles, hip adductors and rectus abdominis — lack of targeted tension, and disrupted breathing.6 These findings come from a pregnant population, so they may not transfer exactly to all women.
Recruiting the wrong muscles
Buttocks, inner thighs and the outer abdominal muscles are the documented substitutes. Visible gripping of the buttocks or thighs usually means those muscles are doing the work. Some gentle deep abdominal activity is normal and expected.
Holding your breath
Disrupted breathing was among the documented errors in the study above. Breathing normally throughout is the standard instruction.
Bearing down instead of lifting
In the 1991 study of 47 women, 25% performed an action the researchers judged could be detrimental to continence status rather than helpful.5 This is the error the mirror check is most useful for catching.
Not releasing fully between contractions
NICE defines pelvic floor training as exercise to improve "strength, endurance, power, relaxation, or a combination of these," and recommends tailoring programmes to a woman's ability to both contract and relax.1 Relaxation is part of the exercise, not a pause between repetitions.
Training only slowly
Long holds and quick contractions train different properties. Most published protocols include both.
The protocol clinical guidelines recommend
Well established NICE recommends a supervised programme of at least eight contractions performed three times per day, for a minimum of three months, as first-line treatment for women with stress or mixed urinary incontinence.17
The word "supervised" carries weight. NICE recommends that programmes be overseen by a physiotherapist or other appropriately trained professional, with at least one review during the programme and one at the end, and that the programme be tailored to the individual woman — her ability to contract and relax, any discomfort, and her own goals.1
What follows is that guideline minimum, described in practical terms. It is a starting point for a conversation with a clinician, not a substitute for one.
Strength and endurance
Lift inward and upward. Hold for as long as is comfortable, then release completely and rest for a similar length of time.
Published protocols commonly build towards holds of several seconds and 8 to 10 repetitions. Start at whatever you can currently manage — if that is two seconds and four repetitions, that is your starting point.
Speed of response
Lift quickly and strongly, then release immediately. Commonly 8 to 10 repetitions.
Early sessions with short holds take a couple of minutes. Once you are holding for longer, the full daily programme is roughly 12 to 18 minutes spread across three sessions. NICE recommends continuing for at least three months before assessing results, and continuing beyond that if it is helping.
The guideline figure is a minimum, but substantially exceeding it has not been shown to produce faster results, and the pelvic floor requires recovery as other muscles do. If you are considering a heavier programme than the guideline minimum, that is a question for a physiotherapist rather than something to decide alone.
Progressing through positions
Limited evidence Progressing from lying to sitting to standing to movement is common clinical practice, and each position places more demand on the pelvic floor than the last. However, no guideline or trial located specifies how long to spend at each stage.
The sequence is generally:
- Lying down, knees bent — the least demanding position in which to learn the movement
- Sitting upright, feet flat
- Standing — full gravitational load, and the position in which most real-world leakage occurs
- During movement — walking, lifting, standing up from a chair
A reasonable approach is to advance when the current position feels controlled rather than on a fixed schedule. A physiotherapist can pace this for your situation, which is precisely the kind of individualisation NICE recommends.
We have deliberately not given week numbers here. Timelines of that kind appear widely online and, as far as we could establish, are not drawn from published evidence.
The knack
Context-dependent "The knack" means contracting the pelvic floor deliberately just before a moment of raised abdominal pressure — before a cough, sneeze or lift.
The original evidence is reasonably strong for its specific context. In a prospective randomised single-blind study of older women with mild-to-moderate stress incontinence, women taught the manoeuvre showed 98.2% less leakage on a medium cough and 73.3% less on a deep cough, measured one week after instruction.8 A later confirmatory study found 76.6% of non-pregnant stress-incontinent women reduced leakage using it, with 18.8% eliminating leakage during the test.9
Two important limitations:
- This evidence is for stress incontinence, and largely for mild to moderate severity. It does not establish benefit for urgency incontinence.
- A 2023 prospective cohort study found that only about one in four women taught the knack actually adopted the pattern while coughing, and noted a lack of evidence for its effectiveness when delivered as a component of a broader training programme.10
So: a measured effect under test conditions, with real uncertainty about how many women carry it into daily life. It may be worth practising, and it is not a guaranteed solution.
When training may not be right for you
This section matters more than any other in this article, and it is frequently omitted.
A pelvic floor can be dysfunctional because it is weak — or because it is over-tense. Strengthening exercises may not be appropriate in the second case, and NICE's own definition of pelvic floor training includes relaxation as a legitimate goal of treatment.1
Stop and seek assessment if you have any of these
- Pelvic pain, or pain during intercourse
- Difficulty starting the urine flow, or a slow stream
- Difficulty emptying the bladder fully
- Constipation or difficulty emptying the bowel
- Pain while doing the exercises
- Symptoms that get worse rather than better after starting
These may indicate that strengthening is not the right approach for you. Continuing regardless is not advisable — see a pelvic floor physiotherapist or your doctor.
Seek medical attention promptly, not eventually
Blood in the urine, pain or burning when passing urine, fever with urinary symptoms, sudden onset of incontinence, or a bulge or heaviness in the vagina all warrant medical assessment rather than an exercise programme.
What to expect, and what nobody can tell you
Well established The one timeline anchored in guidance is this: NICE recommends a minimum programme of three months before assessing results, with review during and at the end.1
Limited evidence Beyond that, individual timelines vary and cannot be reliably predicted. Trials differ considerably in duration, intensity, supervision and how outcomes were measured.
If you encounter an article promising that you will notice a change in week six and see clear improvement by week twelve, treat it with scepticism. We could not find evidence supporting schedules of that kind, and an earlier version of this article contained one that we have since removed.
What can reasonably be said: many women who complete a supervised three-month programme report improvement, a smaller proportion report resolution, and some see no change and go on to discuss other options with a clinician. Which of those applies to you is not predictable in advance.
The adherence problem
Context-dependent When NICE reviewed the evidence for pelvic floor training in prevention, it noted that all the studies showed adherence decreasing over time, and concluded that low long-term adherence is likely to explain the limited evidence for long-term effectiveness.1 Only two of those studies had long follow-up — 12 months in one, 8 years in the other.
That is a narrower finding than "people always stop," but it points at something real: sustaining the programme is a recognised practical difficulty, and NICE made a specific recommendation about encouraging women to continue.
Approaches that may help — none of these are backed by trial evidence specific to pelvic floor training, and they are offered as practical suggestions rather than clinical recommendations:
- Attaching sessions to existing daily routines rather than relying on intention
- Starting at a genuinely easy level in the first week to establish the habit
- Keeping a simple record, so that if you reach the three-month review without improvement you know whether the programme was actually completed
When to seek professional help
NICE recommends supervised training as standard rather than as a fallback, so seeking assessment is not an escalation — it is the recommended route.1
It is particularly worth pursuing if:
- You are not confident you have located the correct muscles — self-assessment is unreliable, and in one study 78% of women contracting incorrectly only corrected it after instruction4
- Three months of consistent training has produced no change
- You have any of the red-flag symptoms listed above
- You do not know which type of incontinence you have
For women unable to perform an effective contraction, NICE suggests considering biofeedback, electrical stimulation or vaginal cones as supplements to training.1 These require professional assessment.
Common questions
Can I do these while driving or at my desk?
Sitting is one of the standard training positions, and attaching the habit to an existing routine is a common practical suggestion. Learning the movement correctly first, ideally with assessment, matters more than where you eventually practise it.
Is it too late if I have been leaking for years?
The Cochrane trials included women with established symptoms, and duration of symptoms is not generally treated as a barrier to trying training. We could not locate reliable evidence on whether outcomes differ by how long symptoms have been present, so we would not claim either that it makes no difference or that starting earlier works better.
Do I need to keep doing them indefinitely?
NICE recommends encouraging women to continue pelvic floor training, and notes that adherence declining over time is associated with the weak long-term evidence.1 Discuss a maintenance approach with your physiotherapist at the end-of-programme review.
Will an app help?
Apps may support adherence, which is a recognised weak point. No app can verify that your technique is correct, which is the reason guidelines recommend supervision.
Should I do these if I am not leaking?
NICE encourages women of all ages to do pelvic floor muscle training and explains that it helps prevent symptoms of pelvic floor dysfunction.1
Can I do them during pregnancy?
NICE recommends a supervised programme of at least three months for pregnant women with stress or mixed urinary incontinence, and considering one from week 20 for pregnant women with a first-degree relative with pelvic floor dysfunction.1 Discuss it with your midwife or maternity team.
Absorbent underwear
We make washable, reusable absorbent underwear for women managing urinary leakage. Elva High 90 has a stated capacity of 90 ml; Elva High 350 has a stated capacity of 350 ml. Our absorbency guide sets out the levels.
Absorbent products manage leakage. They are containment, not treatment, and they do not address the underlying cause. Guidelines position pelvic floor training and clinical assessment as the treatment pathway.
Selected products in our range are CE-marked as medical devices under EU regulation MDR 2017/745. CE marking is a regulatory conformity status covering safety and performance requirements — it is not evidence of clinical effectiveness and does not indicate superiority over other products. We make no claim that our products treat, improve or cure incontinence.
Sources
- NICE guideline NG210, Pelvic floor dysfunction: prevention and non-surgical management, December 2021. Covers first-line status, supervision and review, tailoring to contraction and relaxation, the definition of PFMT, adherence findings in prevention studies, supplementary options, pregnancy recommendations, and prevention advice.
- Dumoulin C, Cacciari LP, Hay-Smith EJC. Pelvic floor muscle training versus no treatment, or inactive control treatments, for urinary incontinence in women. Cochrane Database of Systematic Reviews, 2018. 31 trials, 1,817 women, 14 countries. The 74% vs 11% cure-or-improvement figure derives from 3 trials of 242 women with stress UI; moderate-quality evidence; most trials followed up under 12 months.
- Systematic review of cure rates in incontinence (2017). Median cure rate for surgical interventions in stress UI 82.3% (IQR 72–89.5%); pelvic floor muscle training reported lower cure rates. Cited here only to qualify "first-line," not as a treatment recommendation.
- Cross-sectional study of 779 women in community-based primary care practices. Correct PFM contraction on first attempt: 85.5% (POP), 83.4% (SUI), 68.6% (both), 85.8% (neither). Of 120 women initially contracting incorrectly, 94 (78%) learned after brief instruction.
- Bump RC, Hurt WG, Fantl JA, Wyman JF. Assessment of Kegel pelvic muscle exercise performance after brief verbal instruction. Am J Obstet Gynecol, 1991;165:322–327. n=47, urodynamics assessment. 49% performed an ideal contraction; 25% performed an action judged potentially detrimental to continence. Small, selected, and over thirty years old — included for the error-type finding rather than as a prevalence estimate.
- pelviTrust trial: Influence of a Structured Teaching on Targeted Pelvic Floor Muscle Contraction Ability in Pregnant Women. Healthcare, 2026. n=113 pregnant women. 68% demonstrated correct contraction on visual inspection initially, rising to 97.3% after instruction and verbal feedback. Documented errors: co-contraction of gluteal, adductor and rectus abdominis muscles, lack of targeted tension, disturbed breathing. Pregnant population — may not generalise.
- NICE Quality Standard QS77, Urinary incontinence in women — supervised PFMT of at least three months as first-line treatment; programmes of at least eight contractions three times daily.
- Miller JM, Ashton-Miller JA, DeLancey JOL. A pelvic muscle precontraction can reduce cough-related urine loss in selected women with mild SUI. J Am Geriatr Soc, 1998;46(7):870–874. Prospective, randomised, single-blind. Older women with mild-to-moderate SUI. 98.2% reduction in leakage on medium cough, 73.3% on deep cough, at one week.
- Miller JM et al. Clarification and confirmation of the Knack manoeuvre. 2008. 76.6% of non-pregnant stress-incontinent women reduced leakage using the manoeuvre; 18.8% eliminated it under test conditions.
- Antônio et al. One in four women with stress urinary incontinence who are taught "the knack" manoeuvre adopt this motor pattern while coughing. Neurourology and Urodynamics, 2023. Prospective interventional cohort. Notes a lack of evidence supporting the knack's effectiveness as a component of a PFMT programme.
This article is general information and does not replace individual medical assessment or advice. It describes what clinical guidelines and published research report; it does not tell you what is right for your situation. Clinical guidance recommends that pelvic floor muscle training be supervised and individualised by a qualified professional. If exercises cause pain, if your symptoms worsen, or if you have any of the red-flag symptoms described above, speak to your doctor.
