Grip strength test: how your grip compares for your age

The test

Squeeze a hand dynamometer as hard as you can, sitting with your forearms supported, three times with each hand, alternating. The best of the six readings is your score.

Sources 1, 2

British norms from 49,964 people

Centiles by age and sex from 12 British studies. The middle value peaked at 51 kg for men aged 29 to 39 and 31 kg for women aged 26 to 42.

Source 3

What counts as low

Under 27 kg for men or 16 kg for women is low grip strength in the European sarcopenia group's consensus: a reason for your GP to check further, not a diagnosis.

Source 4

A strong link with health

Each 5 kg lower grip went with a 16 to 20% higher risk of death in large studies in 17 countries and in the UK. That's an association: no trial has shown that training your grip changes it.

Sources 5, 6

Use the same device

Dynamometers can read a few kilograms apart for one person, so test on the same one each time.

Sources 7, 8

Not with a dead hang

We found no study that tested hang time against a dynamometer, and no adult norms for hangs.

Sources 9, 10

Men

The 10th, 25th, 50th, 75th and 90th centiles, in kg:

• Age 20: 30, 35, 40, 46, 52
• Age 25: 36, 41, 48, 55, 61
• Age 30: 38, 44, 51, 58, 64
• Age 35: 39, 45, 51, 58, 64
• Age 40: 38, 44, 50, 57, 63
• Age 45: 36, 42, 49, 56, 61
• Age 50: 35, 41, 48, 54, 60
• Age 55: 34, 40, 47, 53, 59
• Age 60: 33, 39, 45, 51, 56
• Age 65: 31, 37, 43, 48, 53
• Age 70: 29, 34, 39, 44, 49
• Age 75: 26, 31, 35, 41, 45
• Age 80: 23, 27, 32, 37, 42
• Age 85: 19, 24, 29, 33, 38
• Age 90: 16, 20, 25, 29, 33

Source 3

Women

The 10th, 25th, 50th, 75th and 90th centiles, in kg:

• Age 20: 21, 24, 28, 32, 36
• Age 25: 23, 26, 30, 35, 38
• Age 30: 24, 27, 31, 35, 39
• Age 35: 23, 27, 31, 35, 39
• Age 40: 23, 27, 31, 35, 39
• Age 45: 22, 26, 30, 34, 38
• Age 50: 21, 25, 29, 33, 37
• Age 55: 19, 23, 28, 32, 35
• Age 60: 18, 22, 27, 31, 34
• Age 65: 17, 21, 25, 29, 33
• Age 70: 16, 20, 24, 27, 31
• Age 75: 14, 18, 21, 25, 28
• Age 80: 13, 16, 19, 23, 26
• Age 85: 11, 14, 17, 20, 23
• Age 90: 9, 11, 14, 17, 20

Source 3

Who was measured

60,803 measurements of 49,964 people (26,687 of them women) aged 4 to 90, from 12 general population studies in Great Britain measured between 1990 and 2012. Each age's centiles come from a model fitted to all the data, and the people behind each row are those within 2.5 years of that age. The score was each person's best reading, usually of six squeezes.

The authors warn that these centiles come from different people at different ages, so they shouldn't be used to track one person's decline.

Source 3

Why this table

It's large, British and read in full, and it's the source of the European sarcopenia cut-offs and our hand guide's figures (editorial). Across 60 studies, grip norms from other developed regions sat close to these British centiles, while norms from developing regions were lower, so they fit best in countries like the UK.

Sources 3, 4, 11, 12

The nearest age, and the edges between rows

The calculator uses the nearest published age: 47 uses the row for 45, and 22 the row for 20 (editorial). Near the edge between two rows, a year can move you a band: 40 kg is between the 50th and 75th centiles for a man of 22, but between the 10th and 25th for a man of 23, who uses the row for 25 (our arithmetic). The oldest row is 90, so over 92 there are no centiles.

Source 3

Low grip, in numbers

The cut-offs of the European Working Group on Sarcopenia in Older People's 2019 consensus (EWGSOP2), under 27 kg for men and 16 kg for women, are 2.5 standard deviations below the young-adult average in these data: 51.9 kg (SD 9.9) for men and 31.4 kg (SD 6.1) for women, both at age 32. By age 80, 23% of men and 27% of women were at or below them.

Sources 3, 4

Worked examples (our arithmetic)

As the calculator works them out:

• A woman of 47 with 28 kg uses the row for 45: between the 25th (26 kg) and 50th (30 kg) centile, and not low.
• A man of 72 with 25 kg uses the row for 70: below the 10th centile (29 kg), and under 27 kg, so low by EWGSOP2's cut-off.
• A man of 33 with 56 kg uses the row for 35: between the 50th (51 kg) and 75th (58 kg) centile.
• A woman of 85 with 16 kg is between the 25th (14 kg) and 50th (17 kg) centile and not low; at 15.9 kg she's in the same band but under the cut-off, because the centile and the cut-off answer different questions.
• A reading of 66 lb is 29.9 kg, at 0.45359237 kg to the pound, shown to 0.1 kg.

Sources 3, 4

How much a reading moves

In older adults, repeat tests agreed well, with intraclass correlations (a measure of agreement where 1 is perfect) above 0.80 in all but 3 studies, of people with severe dementia, but the smallest real change ranged from 14.5% to 98.5% of the score. In 21 older adults retested after 12 weeks, the typical error was about 1.6 to 2.2 kg (our arithmetic from newtons), and across studies of healthy people the pooled intraclass correlation was 0.92. Grip can differ by a couple of kilograms from day to day.

Sources 13, 14, 15

A marker of overall strengthStrong evidence

Grip "correlates moderately with strength in other body compartments", which is why EWGSOP2 uses it to measure muscle strength, with the chair stand as the alternative for the legs.

Source 4

Grip and death: a strong associationStrong evidence

In 139,691 adults aged 35 to 70 in 17 countries, in the PURE (Prospective Urban Rural Epidemiology) study, each 5 kg lower grip went with a 16% higher risk of death over about 4 years, and grip predicted death better than systolic blood pressure. In 502,293 UK Biobank adults aged 40 to 69, the figures were 20% for women and 16% for men.

Sources 5, 6

Cause and effect isn't shownStrong evidence

The UK Biobank authors note that reverse causation is possible in any observational study and that the links may not mean cause and effect, and the PURE authors call for trials of whether improving strength lowers death rates. We found no such trial.

Sources 6, 5

Exercise raises grip a littleStrong evidence

Across 24 trials in 3,018 older adults, exercise training raised grip strength by a small amount (a standardised mean difference of 0.28), and training the task itself, or mixing types of training, seemed to work. The UK's National Health Service (NHS) asks adults to do strengthening activities on at least 2 days a week.

Sources 16, 17

Low grip makes sarcopenia probable, not certainStrong evidence

In EWGSOP2's words, "sarcopenia is probable when low muscle strength is detected". A diagnosis needs low muscle quantity or quality as well, which takes further tests.

Source 4

Grip and fallsEmerging evidence

The PURE study found no link between grip and injuries from falls, or fractures.

Source 5

Norms differ by regionStrong evidence

Across 60 studies, grip norms from developing regions were lower than the British centiles, and the PURE authors advise reading a grip result against "region/ethnic-specific reference ranges".

Sources 11, 12

What the norms used

Seven of the 12 British studies used the hydraulic Jamar, two the Smedley, a spring dynamometer, two the Nottingham electronic and one the Takei; one of the Jamar studies used the Harpenden at its first wave. Centiles differed by under 10% between devices, and between sitting and standing.

Source 3

Hydraulic, such as a Jamar

EWGSOP2 calls the Jamar "validated and widely used", and says accurate grip testing needs "a calibrated handheld dynamometer under well-defined test conditions".

Source 4

Low-cost digital dynamometers

A low-cost digital Camry read close to a hydraulic Jamar on average: 0.5 kg lower in men and 0.6 kg lower in women aged over 50, and 1.9 kg apart on average in hospital patients. For one person, though, its readings against a digital Jamar Plus differed by up to about 4 to 6 kg.

Sources 18, 19, 8, 20

Other devices read differently

In outpatients aged 60 and over, a Jamar Plus and a Takei read higher than a hydraulic Jamar, which changed who counted as weak. Use the same device each time you test.

Source 7

Why charts differ: one example

Canada's health survey tested healthy adults standing, with a Smedley and the best of two squeezes per hand. Its medians for ages 30 to 34 were 46.9 kg for men and 28.1 kg for women, 3 to 4 kg below the British medians at 30 (our arithmetic). Device, position and country all change a "percentile", so we don't mix tables; the American Society of Hand Therapists' position, seated with the elbow bent to 90 degrees, differs again.

Sources 21, 3, 22

No dynamometer?

The norms came only from dynamometers, so the calculator needs a dynamometer reading (editorial). Squeezing other objects, or timing a hang from a bar, can't be scored against them.

Source 3

A hand or wrist injury, recent hand surgery, or hand pain in the past week

A maximal squeeze loads it: one 2026 study left out people with "hand pain during the evaluation or in the previous week". Wait until it settles, and see our wrist pain, hand arthritis and carpal tunnel guides.

Source 23

A hot, swollen finger joint, or a finger that catches

Rest it, and leave hard squeezing until the finger stops catching, as our ball squeeze move advises. Our trigger finger guide covers a catching finger.

A hand you can't grip with

When hand problems such as arthritis or a stroke make grip testing impossible, EWGSOP2 suggests measuring leg strength instead, and its other strength test is the chair stand. Try our 30-second chair stand test.

Source 4

High blood pressure that isn't yet treated

If your readings are 160/100 or higher and not yet treated, save maximal efforts like this one until treatment has brought your blood pressure under control, as our high blood pressure guide advises (our reading of the guidance). When you do test, breathe out as you squeeze.

Source 21

Diabetic eye disease

With proliferative or severe non-proliferative diabetic retinopathy, or after recent eye laser treatment or surgery, our type 2 diabetes guide advises avoiding hard isometric holds and holding your breath, so ask your eye team before a maximal squeeze.

Under 18, or over 92

This page is for adults, and the British tables stop at 90, so over 92 the calculator shows your reading without a centile.

Source 3

Stop for dizziness or pain

Stop if you feel dizzy or faint, or your hand or wrist hurts, and don't test again that day (our suggestion; the PAR-Q+, the Physical Activity Readiness Questionnaire for Everyone, asks about dizziness).

Source 24

Contact NHS 111 about a wrist injury

Contact the UK's National Health Service (NHS) on 111 for severe wrist pain, a wrist you can't move or hold things with, or a snap, grinding or popping noise at the time of the injury.

Source 25

See a GP about your wrist

If wrist pain stops you doing your normal activities, hasn't improved after 2 weeks, or comes with tingling or loss of feeling.

Source 25

Call 999 (911 in the US) for chest pain

For sudden chest pain that doesn't go away, pain that spreads to your arms, neck, jaw, stomach or back, or chest pain with sweating, feeling sick, light-headedness or shortness of breath.

Source 26

"Squeeze a grip trainer to live longer"Strong evidence

The link between grip and survival is an association, and no trial has tested whether raising your grip lowers your risk. Exercise raises grip a little.

Sources 5, 6, 16

"A dead hang tells you your grip strength"Mixed evidence

No study we found has checked hang time against a dynamometer, and there are no adult norms for hangs. In a bar hang you hold your whole body weight, and in climbing studies bar hangs behaved as tests of shoulder and arm endurance; a finger hang on a small edge is a different test, covered on our climbing page.

Sources 9, 10

"Low grip means you have sarcopenia"Strong evidence

It makes sarcopenia "probable"; a diagnosis needs low muscle quantity or quality as well.

Source 4

"Every grip device gives the same number"Emerging evidence

For one person, devices can differ by a few kilograms, and the choice of device changed who counted as weak.

Sources 7, 8

"Grip norms are the same everywhere"Strong evidence

The British centiles fit other developed regions, but norms from developing regions are lower.

Sources 11, 12