Hamstring strain

Key points

A 2022 guideline on hamstring strain injury in athletes covers sports-related overloading and overstretching injuries in any combination of the three hamstrings, the long muscles at the back of the thigh¹.
Across 36 European men’s professional football clubs followed from 2001 to 2014, 22% of players had one or more hamstring injuries in a season².
A systematic review (a study that gathers the research on a question in a set, checkable way) pooled 63 articles from six field-based team sports³. It put the hamstring injury rate at 0.81 per 1000 hours, and 9.4 times higher in matches than in training³.
In 23 European professional football teams, mean time out rose with the grade on MRI (a scan that shows soft tissue in detail), from 8 days at grade 0 to 73 days at grade 3⁴.
A systematic review of 12 studies found no strong evidence that any MRI finding gives a prognosis for return time after an acute hamstring injury⁵.
In a randomised trial (a study that assigns people to groups by chance) of 75 Swedish elite football players, a protocol emphasising lengthening exercises was followed by a mean 28 days to full team training, against 51 days on conventional exercises⁶.

What it is

What the term covers

A 2022 clinical practice guideline on hamstring strain injury in athletes sets out what it covers¹. A clinical practice guideline is a document of graded recommendations for clinicians, built from a review of the evidence. Its scope is sports-related injuries from overloading and overstretching¹. These injuries affect myofascial structures (muscle and its wrapping) or musculotendinous structures (muscle and its tendon)¹. They can be in any combination of the three hamstring muscles¹.

The three are the biceps femoris, the semitendinosus and the semimembranosus, described on the hamstrings page⁷. A review reports that most hamstring injuries in athletes are strains where muscle joins tendon, most often at the upper junction⁸.

Engraving of the right hip and thigh from behind with the gluteus maximus cut away along its attachments. The hamstrings (biceps femoris, semitendinosus and semimembranosus) run from the ischial tuberosity down the back of the thigh to tendons either side of the knee. — Muscles of the buttock and back of the thighHenry Vandyke Carter, in Henry Gray, Anatomy of the Human Body (1918), Fig. 434. Public domain, via Wikimedia Commons. (source, Public domain)

Two types, by how the injury happens

A case series of professional dancers describes two types of hamstring strain, with different mechanisms⁹. A case series is a report on a group of patients with the same condition. One happens in high-speed running, the other in stretching exercises⁹. A systematic review of 26 studies looked at how each one happens¹⁰:

Running type. The vast majority of studies on running injuries proposed that they occur in the late swing phase of the running stride¹⁰. The swing phase is when the foot is off the ground. Some studies reported the stance phase, when the foot is on the ground¹⁰.
Stretching type. Each study that reported this mechanism concluded that injury occurs with extensive hip flexion (bending at the hip) and a hyperextended knee¹⁰. Hyperextension is straightening a joint past its normal limit.

The same review reports that the running type typically involves the biceps femoris long head and has a shorter recovery time¹⁰. The stretching type commonly affects the semimembranosus¹⁰. The review reports that the running type is the most frequent¹⁰. In Australian football, it reports 81% of hamstring injuries during sprinting and 19% during kicking, a stretching-type injury¹⁰.

How injuries are graded

An international Delphi study asked hamstring injury experts which classification system they used¹¹. A Delphi study is a method that polls a panel of experts in rounds to work toward agreement. The British Athletics Muscle Injury Classification was the most commonly used, at 58%¹¹. The same Delphi study states that such systems lack data on their reliability and validity, and are not specific to single hamstring muscles¹¹.

Who it affects

How often it happens, sport by sport

The rates below come from separate studies, each of one sport or one setting^2,12,13. A cohort study follows a group of people over time, and an athlete-exposure is one athlete taking part in one practice or game.

Who was studied	In matches or competition	In training or practice
36 European men’s professional football clubs, 2001 to 2014	4.77 per 1000 hours	0.51 per 1000 hours²
Professional rugby union, one cohort study	5.6 per 1000 player match hours	0.27 per 1000 player training hours¹²
25 NCAA college sports, 2009-2010 to 2013-2014	5.24 per 10,000 athlete-exposures	2.56 per 10,000 athlete-exposures¹³

In the football cohort, 22% of players had one or more hamstring injuries in a season². In Australian Football League injury data from 1992 to 2012, hamstring strain was the most frequent and prevalent injury¹⁴.

The running page covers 13 international athletics championships¹⁵. In them, the hamstrings were the most frequently injured muscle group in sprints, hurdles, jumps, combined events and men’s middle distances¹⁵.

In NCAA college sports from 2009-2010 to 2013-2014, men had a higher hamstring strain rate than women in soccer, baseball or softball, and indoor track¹³.

Six field team sports, pooled

A systematic review pooled cohort studies from six field-based team sports: soccer, Australian football, rugby union, Gaelic football, hurling and field hockey³. Its 63 articles covered 5952 hamstring injuries and about 7.26 million hours of exposure³. It put the hamstring injury rate at 0.81 per 1000 hours³.

The same review compared rates across those articles³:

Matches and training. The rate was 9.4 times higher in matches than in training³.
Sport. The rate was not significantly different between the six sports³.
Playing surface. The rate was 1.5 times higher on grass than on artificial turf³.
Age. The rate was 6.4% higher for each added year in the average age of the group studied³. The groups’ average body mass and height made no significant difference to it³.

A published correction to the review gives hamstring injuries as 12% of all injuries³. It gives hamstring strain injuries as 10% of all injuries³. The correction says the error it fixed did not affect the rate results³.

Two sources report on change over time^2,3. The pooled review found that the rate had not changed significantly over the last 30 years³. Its authors conclude that more work is needed to reduce it³.

In men’s professional football clubs followed from 2001 to 2014, the training hamstring injury rate rose by 4.0% a year². The total rate rose by 2.3% a year². The authors of that study report that match rates remained stable².

What goes with a higher risk

A meta-analysis (a study that pools the results of several earlier studies) of 78 prospective studies covered 71,324 athletes¹⁶. A prospective study follows people forward in time. Older age and an earlier hamstring strain were the strongest risk factors, as the hamstrings page sets out¹⁶. In the same analysis, a past anterior cruciate ligament injury and a past calf strain were also significant risk factors¹⁶. The anterior cruciate ligament is the front one of two ligaments that cross inside the knee.

Three studies of elite and professional players set muscle measures against hamstring strain injury^17,18,19:

152 elite soccer players. Biceps femoris long head fascicles under 10.56 cm at the start of preseason went with a relative risk of 4.1 for later injury¹⁷. Fascicles are bundles of muscle fibres.
210 elite Australian footballers. Eccentric hamstring strength in the Nordic exercise below 256 N at the start of preseason was associated with a 2.7-fold risk of later injury¹⁸. Eccentric work is muscle work done while the muscle lengthens.
614 professional soccer players. Differences in isokinetic strength between injured and uninjured players were small¹⁹. Isokinetic describes a strength test on a machine that holds a limb to a fixed speed. The authors concluded that strength deficits were weak risk factors that did not identify individual players at risk¹⁹.

What studies found

How long athletes were out

In 23 European professional football teams followed from 2007 to 2011, time out rose with the MRI grade⁴:

Mean days out. Bars start at zero; the table below gives each value with its source.

MRI grade	Mean days out
Grade 0	8⁴
Grade 1	17⁴
Grade 2	22⁴
Grade 3	73⁴

In the same teams, 70% of hamstring injuries were grade 0 or 1, with no sign of fibre disruption on MRI⁴. Those two grades accounted for 56% of the total days out⁴.

Two studies measured something different: the time to return to the level before the injury^9,20. In 18 elite sprinters with a first acute hamstring strain, the median (middle value) was 16 weeks, with a range of 6 to 50²⁰. In 15 professional dancers with a first acute strain from stretching, it was 50 weeks, with a range of 30 to 76⁹.

What scans and examination could tell

A prospective cohort of 180 male athletes with acute hamstring injuries tested how well examination and MRI findings explained return time²¹. A model built from history and clinical examination explained 29% of the variation in time to return to sport²¹. Adding MRI findings explained a further 2.8%²¹. The authors concluded that clinicians cannot give an accurate return time just after the injury²¹.

Two systematic reviews looked at what predicts return time^5,22. One, of 12 studies, found no strong evidence that any MRI finding gives a prognosis for time to return to play⁵. The other, of 16 studies, found no strong evidence for a valuable prognosis from any clinical finding at the start²².

Two studies took scans at the time of return^23,24. In 53 athletes cleared to return to play after a hamstring injury, 89% still showed increased signal within the muscle on MRI²³. A randomised trial compared two exercise programmes after acute hamstring strain injury, and 29 people began a programme²⁴. No participant showed complete resolution of the injury on MRI at return to sport, although pain had nearly resolved and strength had returned²⁴.

When the injury came back

Two sports reported how often the injury was a repeat^12,14:

Professional rugby union. In one cohort study, 23% of hamstring muscle injuries were recurrent, and they cost more days (25) than new injuries (14)¹².
Australian Football League, 1992 to 2012. The recurrence rate was 26%¹⁴.

A case series followed 19 athletes with a hamstring reinjury confirmed on MRI²⁵. In it, 79% of reinjuries were at the same spot in the muscle as the first injury²⁵. More than half happened within 25 days of return to sport²⁵. A systematic review of 11 studies found no strong evidence that any MRI finding predicts reinjury risk²⁶. It found moderate evidence that injuries within the tendon and biceps femoris injuries were associated with a higher risk²⁶.

Trials of exercise programmes after a strain

Four randomised trials compared exercise programmes in athletes with an acute hamstring injury^6,27,28,29:

Who was studied	What was compared	Time to return	Repeat injuries
75 Swedish elite football players	A protocol emphasising lengthening exercises against conventional exercises	Mean 28 against 51 days to full team training	One, in the conventional group⁶
56 Swedish elite sprinters and jumpers	A protocol emphasising lengthening exercises against conventional exercises	Mean 49 against 86 days to full training	Two, both in the conventional group²⁷
90 male athletes, injury confirmed on MRI	Lengthening exercises from day 1 against after running at 70% of maximal speed	Median 23 against 33 days; no significant difference	No significant difference in rates over 12 months²⁸
24 athletes	Stretching and isolated hamstring strengthening against progressive agility and trunk stabilisation	Mean 37.4 against 22.2 days; no significant difference	7 of 10 against 1 of 13 within a year of return²⁹

The authors of the 90-athlete trial reported a greater than expected loss to follow-up, at 19%²⁸. They said the trial was potentially underpowered for a firm conclusion²⁸.

Two reviews pooled trials like these^30,31. A Cochrane review (a systematic review done to Cochrane’s published standards) found two randomised trials with 104 participants³⁰. It concluded that current practice and widely published protocols could be neither supported nor refuted³⁰. A meta-analysis of 10 randomised trials pooled two studies of lengthening exercises³¹. It found a significantly shorter time to return to play but no difference in the risk of re-injury³¹.

Trials of platelet-rich plasma injections

Platelet-rich plasma is a part of a person’s own blood with extra platelets, prepared for injection. Trials and pooled analyses of it disagree^31,32,33,34:

A randomised trial with placebo injections, in 80 competitive and recreational athletes. The plasma injections showed no benefit over the placebo (dummy) ones in time to return to play or re-injury rate at 1 year³².
A randomised three-arm trial of 90 professional athletes. Return to sport came 5.7 days sooner after one platelet-rich injection than after a platelet-poor one³³. Against no injection, the 2.9-day gap was not significant³³.
A meta-analysis of three studies. It found no effect on time to return to play compared with control³¹.
A 2026 meta-analysis of six randomised trials with 277 participants. Return to play came a mean 8.6 days sooner with the injections, with high heterogeneity, meaning the trials’ results varied widely³⁴. Reinjury did not differ, and the certainty of evidence was rated low to moderate³⁴.

Programmes in athletes without an injury

Reviews of programmes meant to lower the rate of hamstring injury do not agree^35,36,37. Two of the three below concern the Nordic hamstring exercise, set out on the Nordic hamstring curl page^36,37.

A Cochrane review of seven randomised trials, with searches to December 2008. It concluded there was insufficient evidence from randomised trials to draw conclusions for football and other high-risk activities³⁵.
A meta-analysis of 15 studies of athletes across sports. Programmes that included the Nordic hamstring exercise had a pooled hamstring injury risk ratio of 0.49, against programmes or usual training without it³⁶. A risk ratio is how many times more often something happens in one group than in another.
A reanalysis that kept randomised trials alone. Its pooled risk ratio was 0.59, with a 95% confidence interval of 0.27 to 1.29³⁷. A confidence interval is a range around a result that shows how precise it is. The authors judged the protective effect inconclusive and mostly based on trials at high risk of bias³⁷.

Surgery when the tendon pulls off the bone

An avulsion is a tear in which a tendon pulls away from the bone it attaches to. The proximal hamstring tendon page cites a systematic review on how common this is³⁸. It reports that avulsion of the hamstring origin from the ischial tuberosity (the “sitting bone”) makes up 3% to 11% of hamstring injuries³⁸.

A randomised trial enrolled 119 patients aged 30 to 70 with a proximal hamstring avulsion³⁹. Function scores at 2 years were similar after surgical reinsertion of the tendons and without surgery, 79.9 against 78.5 out of 100³⁹. The approach without surgery was noninferior, meaning not meaningfully worse beyond a margin set in advance³⁹.

Connections

The injury sits in the hamstrings, and one of the three muscles carries most of it in football⁴⁰. In male professional footballers, 80% of hamstring injuries placed in one muscle from 2011/12 onward were in the biceps femoris⁴⁰. Avulsions of the upper tendon are covered on the proximal hamstring tendon page³⁸.

The sciatic nerve is the large nerve that runs from the lower back down the back of the thigh. One study read the MRI scans of 96 hamstring strain injuries in US college American football players⁴¹. Swelling touched or surrounded the sciatic nerve in 75 of them, or 78.1%⁴¹.

Related exercise and stretch pages: Nordic hamstring curl, running and hamstring stretch^15,36,42. A four-season study followed elite soccer teams from Iceland and Norway, which chose whether to use the programmes on offer⁴². The incidence of hamstring strains did not differ between teams that used a flexibility training programme and teams that did not⁴².