Goals

Splits by age: what changes, what does not

Measured joint range declines by a handful of degrees across five decades, so age is a far smaller factor than the question assumes. What changes is your starting point, your recovery and the time you have. Hip bone shape is the real ceiling, and it was there at eighteen.

29 Jul 2026 · Revised 10 Aug 2026 · 8 min read

The question behind splits by age is usually "am I too late", and the honest answer is that age is a smaller factor than almost everybody assumes and a real one. The population data shows a decline of a handful of degrees across five decades — not the collapse the question implies.

What changes with age is mostly your starting point, your recovery, and the number of years you have been carrying whatever you carry. None of those is a ceiling. Your hip anatomy might be, and it was there at twenty as well.

The short answer

  • Measured joint range declines by a few degrees over five decades, not by half.
  • Cross-sectional data overstates even that, because it compares different people rather than following the same ones.
  • Adults of every age improve their range with training. That is well supported.
  • No trial has studied the splits at any age. Everything specific to the position is anecdote.
  • Hip bone shape is the real ceiling, and it does not arrive with age — it was always there.

Splits by age

Start with what the population measurements actually show.

Roach and Miles analysed joint range across adulthood using NHANES I data in Physical Therapy in 1991 — a genuine population sample rather than a convenience group of volunteers. Across five decades, the decline in most joint ranges was on the order of a few degrees, with hip motion among the more affected. It is a real decline and it is far smaller than the folk model of stiffening.

Soucie and colleagues published reference values for joint range in 674 healthy people aged 2 to 69 in Haemophilia in 2011, and the pattern is the same shape: range is greatest in childhood, declines through adolescence into adulthood, and then changes slowly.

Grade: reasonably solid, and cross-sectional. These studies compare a 25-year-old to a 55-year-old rather than following one person for thirty years, so they mix ageing with cohort differences in occupation, activity and body size. Cross-sections tend to overstate age effects for exactly that reason.

Joint laxity follows a similar curve. Beighton and colleagues, in the Annals of the Rheumatic Diseases in 1973, measured hypermobility across a population and found it commonest in the young and declining with age. And Hakim and colleagues' twin study in Arthritis & Rheumatism in 2004 found joint laxity largely heritable — which is a reminder that the person in the splits photograph may have started with something you did not.

What actually changes

Four things, and only one of them is about tissue.

Your starting point. A 45-year-old who has sat at a desk for twenty years is further from a split than a 25-year-old who has sat for two. That is accumulated exposure, not age.

Recovery between sessions. Older adults generally take longer to recover from hard work, which argues for smaller, more frequent doses rather than heroic sessions. Feland and colleagues found something consistent with this in Physical Therapy in 2001: in people over 65, 60-second holds outperformed 30-second holds, where in younger adults 30 matched 60. The dose that works appears to shift with age.

Tissue properties. Tendon stiffness and collagen cross-linking do change over decades. How much that limits achievable range, as opposed to how it feels, is not well quantified.

Confidence at end range. Which is not trivial. A large part of the acute range anybody has is their nervous system's judgement about how far is safe, and that judgement gets more conservative with a history of injuries.

What does not change with age

Bone. Frank and colleagues reviewed imaging of people without symptoms in Arthroscopy in 2015 and found cam-type hip morphology — extra bone at the femoral head-neck junction — in roughly a third of asymptomatic volunteers and over half of asymptomatic athletes. Socket depth and orientation vary as well.

That ceiling is the one that decides whether a split is available to you, and it was there at eighteen. People who assume age closed the door were often meeting an anatomical limit that had been waiting the whole time. This is more encouraging than it sounds: if you could not do it at twenty-five either, nothing has been lost.

The response to training. Stathokostas and colleagues reviewed flexibility training in older adults in the Journal of Aging Research in 2012 and found range improved with training across the studies. They also noted, and it must be said alongside, that the evidence for carry-over into everyday function was much weaker than the evidence for the range gain itself.

Grade: reasonably solid for the range gain, thin for what the gain buys you.

What nobody has studied

There is no trial of any splits programme, at any age. Not one. Everything published about how long a split takes, at 20 or 50, is coaching tradition and anecdote.

That is not a reason to dismiss the coaching tradition, which is where most useful practice in this area comes from. It is a reason to be suspicious of anybody who states a timeline as though it were a finding, and it is why this cluster prints no date. Front splits progression and middle splits progression both work on that footing, and so does an honest splits timeline, which asks your age band, your gap to the floor and how often you train, then answers in six-week training blocks and a wide range of months rather than a date.

The same applies to our own product. Limber's engine has a splits goal; the eight-week front-splits and ten-week middle-splits tracks are listed as not drawn, because the 56-move catalogue holds three hip flexor moves and the app's ten body areas do not include the adductors at all. Publishing a repeated week and calling it a progression is the refusal.

Training the splits at 40, 50, 60

The adjustments are modest and they follow from the four changes above.

  • Longer holds. Feland's finding suggests 60 seconds rather than 30 over 65, and it is a reasonable default anywhere past middle age.
  • Warm properly, and always. Range is markedly lower cold, and the consequences of a cold maximal position are worse with less tissue tolerance behind them.
  • More frequently, less hard. Total accumulated weekly time is the dominant variable in the dose literature — Thomas and colleagues in the International Journal of Sports Medicine in 2018 — so distribute it.
  • Load the range. Afonso and colleagues' review in Healthcare in 2021 found strength training and stretching produced comparable range gains, on a small and varied evidence base the authors flagged themselves. Grade: mixed — and for anyone whose tissue tolerance is the constraint, adding load has an argument that hanging in a position does not.
  • Stop at pinches. A hard pinch at the front of the hip or deep in the groin is bone or joint contact rather than a stretch. It does not soften with time and it is a stop.

Groin pain that catches or clicks painfully, hip pain persisting after the session, pain radiating down the leg, numbness or weakness needs a clinician who can examine you. That advice does not change with age; the base rate of things worth examining does. Nothing here diagnoses anything and nothing here is treatment.

Measure your own decade

Population averages are a poor guide to a person. The distribution at any age is wide enough that a 55-year-old can sit above the median 30-year-old, and a reading tells you where you are rather than where your cohort is.

The hip flexor test reads hip extension on each side in three bands — 25 above the bench line, 60 level, 90 below. Two honest limits: it is not a valid measure of hip extension unless pelvic tilt is controlled, per Vigotsky and colleagues in PeerJ in 2016, so press the small of your back flat; and the smallest gap it can show between your sides is 30 points against an imbalance threshold of 12, so any difference at all is flagged and a one-band gap means measure again.

The Range Score composes whatever you have measured into a single 0–100 reading. One thing it will not do is tell you how you compare to other people your age: no population has ever been measured against this scale, and the app says so. It exists so the distance between two of your own readings means something, not so you can rank.

Take a baseline, work for six weeks, take it again under the same conditions. Can adults become flexible covers the wider question, and the cluster is goals.

Questions

Can you get the splits at 40?

There is no published data on the splits at any age, so nobody can give you odds. What is established is that adults of every age improve their range with training, and that measured range declines only a few degrees across five decades. Hip bone shape is the real ceiling, and it was there at twenty as well.

Is it too late to become flexible after 50?

No. Reviews of flexibility training in older adults find range improves with training. The same reviews find the evidence for carry-over into everyday function much weaker than the evidence for the range gain, which is worth knowing — it is a good reason to want the range and a poor reason to expect it to change your life.

How much flexibility do you lose with age?

Less than the folk model suggests. Population measurements across five decades show declines on the order of a few degrees for most joints, with hip motion among the more affected. Cross-sectional studies also tend to overstate this, because they compare different people rather than following the same ones.

Should older adults hold stretches longer?

Probably. A trial in people over 65 found 60-second holds outperformed 30-second holds, where in younger adults 30 seconds matched 60. That single finding is the best guide available and it is one study, so treat 60 seconds past middle age as a reasonable default rather than a law.

Why could I do the splits as a child?

Because joint laxity is greatest in childhood and declines through adolescence, and because most children who could do it were practising it regularly. Losing it is a combination of that natural decline and thirty years of not visiting the position — the second of which is the part you can address.

Does age change how fast I improve?

Almost nothing is published on the rate, because hardly any trial has compared age groups on the same protocol. Treat age as a reason to expect a lower starting point, longer recovery and possibly longer holds — not as a reason to expect no change.

Take the reading
Nearby in this cluster
Sources
  1. Roach KE, Miles TP. Normal hip and knee active range of motion: the relationship to age. Physical Therapy 1991;71(9):656–65. doi:10.1093/ptj/71.9.656
  2. Soucie JM, Wang C, Forsyth A, Funk S, Denny M, Roach KE et al. Range of motion measurements: reference values and a database for comparison studies. Haemophilia 2011;17(3):500–7. doi:10.1111/j.1365-2516.2010.02399.x
  3. Beighton P, Solomon L, Soskolne CL. Articular mobility in an African population. Annals of the Rheumatic Diseases 1973;32(5):413–8. doi:10.1136/ard.32.5.413
  4. Hakim AJ, Cherkas LF, Grahame R, Spector TD, MacGregor AJ. The genetic epidemiology of joint hypermobility: a population study of female twins. Arthritis & Rheumatism 2004;50(8):2640–4. doi:10.1002/art.20376
  5. Feland JB, Myrer JW, Schulthies SS, Fellingham GW, Measom GW. The effect of duration of stretching of the hamstring muscle group for increasing range of motion in people aged 65 years or older. Physical Therapy 2001;81(5):1110–7. PMID 11319936
  6. Frank JM, Harris JD, Erickson BJ, Slikker W 3rd, Bush-Joseph CA, Salata MJ et al. Prevalence of Femoroacetabular Impingement Imaging Findings in Asymptomatic Volunteers: A Systematic Review. Arthroscopy 2015;31(6):1199–204. doi:10.1016/j.arthro.2014.11.042
  7. Stathokostas L, Little RM, Vandervoort AA, Paterson DH. Flexibility training and functional ability in older adults: a systematic review. Journal of Aging Research 2012;2012:306818. doi:10.1155/2012/306818
  8. Thomas E, Bianco A, Paoli A, Palma A. The Relation Between Stretching Typology and Stretching Duration: The Effects on Range of Motion. International Journal of Sports Medicine 2018;39(4):243–254. doi:10.1055/s-0044-101146
  9. Afonso J, Ramirez-Campillo R, Moscão J, Rocha T, Zacca R, Martins A et al. Strength Training versus Stretching for Improving Range of Motion: A Systematic Review and Meta-Analysis. Healthcare 2021;9(4). doi:10.3390/healthcare9040427
  10. Vigotsky AD, Lehman GJ, Beardsley C, Contreras B, Chung B, Feser EH. The modified Thomas test is not a valid measure of hip extension unless pelvic tilt is controlled. PeerJ 2016;4:e2325. doi:10.7717/peerj.2325

Each address was followed to the record it names before it was written down. Where a paper could not be resolved that way, the page names it in the prose and leaves it unlinked rather than guessing at an address.

← All of Goals

What a field note is

One question, answered as far as the evidence goes and no further. Where the research is thin the note says so, and where there is none it says that instead of writing around it.

It is not advice and it does not diagnose anything. If pain radiates down a limb, followed an injury, or comes with numbness or weakness, see a clinician rather than a web page.

Reading, then a reading.

Every instrument here scores on the app’s own range_math table, and none of them asks for an account or keeps anything.

Read the reasoning here, take the measurement there, and take it again in three weeks.

The instruments