Sports trauma · Lower limb

Calf muscle injury (triceps surae): tennis leg, soleus and differential diagnosis

A patient arrives with a calf that has been painful since yesterday, swollen, and that they cannot roll through when walking. Before laying hands on it, two questions must have been answered: is it really muscle, and is the Achilles tendon intact? Both answers are obtained in the clinic, in a few minutes, and they decide everything else.

In brief: the clinical synthesis

What this chapter gives: the short version, the one you reread in the corridor before going into the room. The rest of the article details each of these lines and gives the source of every figure.

Calf muscle injury is a common and deceptive presentation. Common: the medial head of gastrocnemius is the third most injured muscle in the elite athlete, after biceps femoris and rectus femoris (Bright 2017, PMID 28661826). Deceptive: behind an identical clinical picture (sudden pain in the back of the leg, swelling, limping), hide at least four entities, two of which are emergencies.

The approach comes down to three steps, in this order, and the order is not negotiable.

First step, rule out deep vein thrombosis. In the reference ultrasound series of 141 patients referred with a clinical diagnosis of tennis leg, 14 of them (9.9 %) had no muscle injury at all: they had an isolated DVT (Delgado 2002, PMID 12091669). One patient in ten. That is why massage, vigorous mobilisation or pneumatic compression applied to an uninvestigated calf are not neutral acts.

Second step, rule out Achilles tendon rupture. The calf squeeze test, known as the Thompson or Simmonds test, has a sensitivity of 0.96 and a specificity of 0.93 in a prospective series of 174 ruptures and 28 unconfirmed suspicions (Maffulli 1998, PMID 9548122). It costs fifteen seconds.

Only then, third step, locate the muscle injury. Medial gastrocnemius at the myotendinous junction, soleus, plantaris: the picture, the timelines and the reinjury risk differ markedly. The soleus dominates heavily in volume: 84.6 % of calf injuries in the elite Australian football registry, and accounts for 91.4 % of reinjuries (Green 2020, PMID 31494970).

On prognosis, one landmark is worth more than any table of timelines: it is the state of the connective tissue, not the volume of muscle oedema, that predicts how long the patient will be unavailable. Across 100 consecutive patients, the mean time to return went from 8 days when the connective interfaces were intact to 48 days when they were disrupted (Prakash 2018, PMID 29074478).

This article covers injury of the triceps surae. For injury from a direct blow, whose pathophysiology and complications are different, see muscle contusion (dead leg) and its complications. For chronic tendon distress in the same muscle-tendon complex, see Achilles tendinopathy.

Why is acute calf pain not always muscular?

This is the chapter that involves patient safety, and that is why it comes first rather than in the middle. A painful, swollen calf looks like a painful, swollen calf, whatever the cause.

Clinical experience builds a mental picture of tennis leg: the tennis or badminton player in their forties, the sensation of being kicked or hit by a stone in the muscle belly, sometimes an audible snap, the immediate inability to push off the forefoot. That picture is accurate, and that is precisely what makes it dangerous: it is so characteristic that it discourages looking for anything else.

The reference series on this point is old but has not been refuted. Delgado and colleagues reviewed the ultrasound scans of 141 patients referred for imaging with a clinical diagnosis of tennis leg already made. In other words, patients whom a clinician had already examined and labelled. This is what ultrasound found.

Two results from this series deserve to be kept separate.

The first is the safety figure: an isolated DVT in 14 patients out of 141, that is roughly one patient in ten, in subjects already examined by a clinician who had concluded to a muscle injury. That is not a statistical curiosity, it is a measured clinical error rate.

The second corrects a stubborn belief. Tennis leg was long taught as a rupture of the plantaris. In this series, the plantaris is at fault in only 2 cases out of 141, that is 1.4 %. The muscle really involved is the medial head of gastrocnemius, in two thirds of cases.

Why the confusion is structural, and not anecdotal

DVT and calf muscle injury share their cardinal signs: calf pain, increased volume, local heat, pain on passive dorsiflexion, discomfort on weight-bearing. They also share part of their background: recent immobilisation, prolonged travel, age. And they can coexist: Slawski reported a case of ipsilateral thrombophlebitis occurring on an already established tennis leg, recalling in passing that the reverse error, taking a tennis leg for thrombophlebitis, had led to unnecessary anticoagulation and to bleeding complications (Slawski 1994, PMID 8027900).

This point matters so as not to swing to the opposite excess. The aim is not to send every painful calf to the vascular emergency department: it is to know how to recognise the situations that cannot be settled in the clinic, and not to treat before they are settled.

The Wells score guides, it does not rule out

The Wells score is the usual landmark for pre-test probability of DVT. It is useful to the physiotherapist for structuring the history, but you have to know what it does not do. In 1,135 inpatients with suspected DVT, the area under the ROC curve for the score was only 0.60, and 5.9 % of patients classed as low probability nevertheless had a proximal DVT: the authors conclude that the score “is not sufficient to rule out DVT or to guide decisions” in that setting (Silveira 2015, PMID 25985219).

The practical consequence is clear and it is liberating: the physiotherapist's role is not to calculate a probability in order to decide whether to treat. It is to spot the doubt and pass it on. The decision to image, to anticoagulate or to monitor belongs to the doctor, and it is itself debated: for symptomatic isolated distal DVT in the low-risk patient, therapeutic anticoagulation has not been shown superior to placebo for reducing thromboembolic events, at the price of increased bleeding risk (Righini & Robert-Ebadi 2024, PMID 38546285). If the question divides haemostasis specialists, it is not settled on a treatment couch.

The differential table

The five entities that make up most acute calf pain, with what separates them.

A remark on the ruptured popliteal cyst: it is rare but it is not negligible in a population presenting with a swollen leg. Across 106 patients referred to vascular surgery for a swollen leg, DVT accounted for 44.8 % of causes and a ruptured Baker's cyst for 2.4 % (Sato 2001, PMID 11381504). The lesson of that series is above all that fewer than one swollen leg in two is a thrombosis : the doubt must be investigated, not turned into certainty in one direction or the other.

How do you rule out an Achilles tendon rupture in the clinic?

Achilles rupture is the other major differential, and it is the easier of the two to rule out: unlike thrombosis, here there are clinical tests whose diagnostic value has been measured, and it is excellent.

The patient who ruptures their Achilles tendon very often describes the same thing as the patient who tears their medial gastrocnemius: a violent blow from behind, a snap, the sense of something having been thrown at their leg. The difference is not in the story, it is in the examination.

Who ruptures their Achilles tendon

Knowing the background helps to calibrate vigilance. A recent meta-analysis of 28 population studies, covering more than 630 million people and 568,000 ruptures, puts the worldwide incidence at 15.7 per 100,000 person-years, rising by a mean 2.7 % a year (95 % CI 2.0-3.3) over six decades. Men are affected 3.18 times more often than women (95 % CI 2.50-4.04), with a peak in men aged 30 to 49 at 42.6 per 100,000 person-years. About 68 % of ruptures are sport-related (Kotsifaki 2026, PMID 41933260).

The four values come from the same source: Kotsifaki R, Malliaras P, Byron C, et al. Sports Med 2026;56(6):1467-87 (PMID 41933260), meta-analysis of 28 population studies, 1950-2022.

In other words, the occasionally sporting man in his forties, exactly the tennis leg profile, is also the Achilles rupture profile. The two populations overlap. That is why the test is not done “if in doubt”: it is done routinely.

The calf squeeze test, known as the Thompson test

Patient prone, feet over the end of the couch, ankles free. The examiner grasps the calf muscle belly and squeezes it transversely. If the muscle-tendon complex is continuous, the squeeze deforms the muscle, pulls on the tendon and produces visible plantarflexion. If the tendon is ruptured, the chain is broken: the foot does not move.

Its performance was measured in a prospective series run over thirteen years, comparing 174 patients with confirmed complete subcutaneous rupture and 28 patients suspected but without rupture (Maffulli 1998, PMID 9548122).

Two practical lessons come out of those figures.

First, palpation alone is not enough. It is the least sensitive test, and it is nevertheless the commonest reflex. A tendon that feels continuous under the fingers is not necessarily so: the haematoma fills the gap, muscle guarding blurs the contour.

Second, and this is the point Maffulli himself stresses, in the 174 genuinely ruptured patients, at least two tests were positive. The clinical strategy that follows is not to choose the best test, but to perform two or three routinely.

The tests to combine

• Calf squeeze (Thompson-Simmonds). Prone, feet over the end of the couch. Absence of plantarflexion on squeezing the triceps = abnormal.
• Matles test. Prone, active flexion of both knees to 90°. Observe the spontaneous position of the feet. On the sound side, the ankle stays in slight equinus under the effect of tendon tone. On the ruptured side, the foot drops into neutral or dorsiflexion.
• Loss of resting tension. Side-to-side comparison of the physiological resting equinus, with the patient completely relaxed.
• Level-by-level palpation of the tendon, from the calcaneus to the myotendinous junction, looking for a gap.
• Rising onto the toes in single-leg stance, when the pain allows. Inability is a strong sign, but its presence is not specific: a painful muscle injury also prevents it.

The combination of signs has been evaluated directly. In 66 patients operated on for an acute rupture confirmed intraoperatively, the triad abnormal Thompson test + loss of resting tension + palpable gap was present in all of them, that is a sensitivity of 100 %. In the same study, MRI read 60 complete ruptures, 4 partial ones and 2 inconclusive examinations: it was therefore less accurate than clinical examination, while delaying management by several days (Garras 2012, PMID 22538958).

A useful point to avoid a classic false negative: a ruptured Achilles tendon does not abolish all active plantarflexion. The toe flexors, tibialis posterior, the peroneals and, when it is intact, the plantaris still produce movement. A patient who “can move the foot” does not therefore have a continuous tendon. This trap is documented right into the case literature (Rohilla 2013, PMID 23345486).

What do you need to know about the anatomy of the triceps surae to understand the injury?

Three anatomical facts explain on their own where the injuries sit, their mechanism and how to rehabilitate them. The rest is descriptive detail that clinical practice does not need.

First fact: the gastrocnemius crosses the knee, the soleus does not. The two heads of gastrocnemius arise above the femoral condyles: they are biarticular, and their tension depends simultaneously on the position of the knee and that of the ankle. The soleus arises below the knee, from the tibia and fibula: it is monoarticular, and only the ankle concerns it. That difference governs the injury mechanism, the clinical test and the strengthening exercise.

Second fact: the injury occurs at the interface, not in the muscle belly. The medial head of gastrocnemius ends on a broad aponeurosis that glides against that of the soleus. It is that transition zone, both the stiffest and the most loaded, that gives way. About two thirds of calf injuries sit at the fascial junction between the medial head of gastrocnemius and the soleus (Bright 2017, PMID 28661826).

Third fact: the plantaris exists, but it explains almost nothing. That small vestigial muscle, absent in part of the population, long carried the blame for tennis leg. Delgado's ultrasound series implicated it in 2 cases out of 141, that is 1.4 % (Delgado 2002, PMID 12091669).

The injury mechanism, deduced from the anatomy

The most dangerous position for the medial head of gastrocnemius combines knee extension and ankle dorsiflexion : the two joints it crosses lengthen it at the same time. That is the forward lunge of the tennis or badminton player, the uphill sprint start, the missed step off a kerb. The muscle is then loaded eccentrically at maximal length, exactly the condition that produces stretch injuries.

For the soleus, the knee protects nothing. It works in every knee position, which explains two things: its far greater cumulative loading during running, and the fact that its injuries appear more readily on a recent training load than on a single movement. It is an overload injury more than an accident injury.

Does the site of the injury really change the prognosis?

Yes, but not in the way it is often put. The difference in timeline between soleus and gastrocnemius exists in the data; it does not reach statistical significance. What does differ clearly is the clinical profile and the reinjury risk.

The Australian Football League soft tissue injury registry is the best source available on this question, because it prospectively follows a homogeneous population with routine MRI. Across 184 calf injuries collected between 2014 and 2017 in 16 clubs (Green 2020, PMID 31494970) :

• The soleus accounts for 84.6 % of injuries. That is probably the most counter-intuitive result in the recent calf literature: the emblematic tennis leg muscle, the medial gastrocnemius, is in the minority in a population of field athletes.
• The mean time to return was 25.4 ± 16.2 days for the soleus against 19.1 ± 14.1 days for the gastrocnemius. Six days' difference, but with P = 0,097, that is a difference that is not significant at the conventional threshold.
• Injuries occurring while running took about 12 days longer to recover than those occurring outside running (P = 0.001), and that difference is solid.
• 91.4 % of reinjuries involved the soleus.

That last point is the most operational. The soleus may not significantly lengthen the time to first return, but it massively dominates relapses. The issue with the soleus is therefore not “how long”, it is “will it hold”.

The gastrocnemius data have been refined recently. Across 82 MRI-confirmed gastrocnemius injuries between 2014 and 2023 in the same registry, the medial head accounts for 78 % of cases, and the median times to recovery are 3 days to walk without pain, 14 days to run at more than 90 % of maximal speed, 14 days to return to full training and 19 days to return to competition. More than one injury in six was a reinjury, and 79 % of those reinjuries occurred within six months (Gray 2025, PMID 40985375).

How to tell soleus from gastrocnemius on examination

The manoeuvre follows directly from the biarticular arrangement. You test resisted plantarflexion in two successive positions:

That reservation deserves to be explicit rather than hidden. A large part of calf semiology rests on coherent anatomical reasoning whose sensitivity nobody has published. Saying so does not weaken the approach: it avoids announcing to the patient a certainty you do not have.

How do you classify the injury, and what should you really expect from imaging?

Three systems coexist, and the clinician does not need to memorise them all. They need to understand what these systems measure, and what they all measure, in the end, is the same thing: involvement of the connective tissue.

For a long time, muscle injuries were graded on the extent of fibre involvement: grade 1 strain, grade 2 partial tear, grade 3 complete rupture. That classification never predicted timelines well, for a reason we understand better today: it is not the muscle that limits return, it is the connective tissue that transmits its force.

What modern grading really measures

The study that established this point most clearly retrospectively assessed 100 consecutive patients referred with suspected calf injury, that is 114 injuries on MRI. The authors graded each injury from 0 to 3 by the involvement of the connective structures and their interfaces, then related that grade to the actual time to return. Connective involvement was present in 63 of the 100 patients, and connective tissue failure, grade 3, in 18 (Prakash 2018, PMID 29074478).

A sixfold difference within the same condition, on a criterion that clinical examination cannot see. It is the strongest argument in favour of imaging when a prognosis has to be announced, to an athlete at the end of a season, to a patient whose job requires standing.

The same mechanism is found in a different population. Across 149 calf injuries in the Australian registry, players with severe aponeurotic involvement took 31.3 ± 12.6 days to return, against 19.4 ± 10.8 days in its absence (P = 0.003) (Green 2020, PMID 33030961).

The three systems in use

The Pedret classification deserves a further word, because it is designed specifically for tennis leg. The five types are: type 1 myo-aponeurotic; type 2A involvement of the gastrocnemius aponeurosis over less than 50 % of its width; type 2B over more than 50 %; type 3 involvement of the free, tendinous aponeurosis; type 4 mixed. The longest timelines are associated with involvement of the free aponeurosis. Two ultrasound signs accompany the forms with a poor prognosis: the presence of an intermuscular haematoma and asynchronous movement between gastrocnemius and soleus during ankle flexion-extension: signs that appear when more than 50 % of the aponeurotic width is involved (Pedret 2020, PMID 32854168).

Should you image, and with what

The honest answer is: most often, no, but when you do image, you know what to look for.

Ultrasound is the first-line investigation. It is available, dynamic, it shows the inter-aponeurotic fluid, and above all it sees deep vein thrombosis, which settles both questions at once. A review of the use of ultrasound in calf injuries concludes that the great majority of injuries can be diagnosed and managed with no further imaging, and that repeated ultrasound allows healing to be followed (Bright 2017, PMID 28661826).

MRI brings the fine connective grading. It is justified when a precise prognosis has a consequence (selection, contract, long sick leave) or when the course does not follow. In the Pezzotta series, the extent of the oedema emerged as an independent prognostic factor for time to return in two multivariable regression models (Pezzotta 2017, PMID 28770309).

That result, from the same Australian cohort, deserves to be taken seriously (Green 2020, PMID 33030961). It means that a fine follow-up MRI does not reassure about the risk of relapse, and that an image that is still imperfect in a patient whose functional criteria are met is not necessarily a reason to delay. The return decision is taken on function, not on the image.

What are the risk factors, and what do we know about reinjury risk?

Few risk factors survive systematic review, and those that do are uncomfortably simple: age, and having already had the same injury. Neither is modifiable.

The reference systematic review retained 10 studies assessing risk factors for calf injury, representing 5,397 athletes and 518 calf or leg injuries, in football, Australian football, rugby union, basketball and triathlon. Its synthesis by level of evidence gives (Green & Pizzari 2017, PMID 28259848) :

Two comments are needed on that figure.

The first concerns the absence of association with sex, height and weight. That is not evidence of no effect: it is the absence of evidence of an effect, in a literature the authors themselves describe as scarce and biased. It would be excessive to conclude that these variables do not matter.

The second is more useful in the consultation. Since the two solid factors are not modifiable, all prevention plays out on what is not in this list : load management, calf capacity, the quality of the initial rehabilitation. That is exactly what the 20 international experts interviewed about their practices report: there is no universal preventive programme for the calf, and strategies are individualised to the athlete's characteristics and the demands of their sport (Green 2022, PMID 35032233).

The reinjury risk, quantified

This is where the calf differs clearly from other muscle injuries, and where the published figures need to be read carefully, because they depend entirely on the definition of “reinjury” adopted.

An analysis of 563 calf injuries in 359 players over ten years measured that definitional effect precisely (Green 2025, PMID 40919406) :

The four values come from the same source: Green B, Schache AG, Pizzari T. BMJ Open Sport Exerc Med 2025;11(3):e002865 (PMID 40919406). Reinjuries cost at least 2,153 cumulative days of unavailability over the decade, that is at least 35.6 days on average per reinjury.

The figure to remember is the third. One subsequent injury in five occurs in a player who had not finished recovering, and it then costs 46.7 days on average, more than double an index injury. It is the quantified demonstration that returning too early does not save time: it costs time.

The clinical factors associated with reinjury have been identified separately. For an early reinjury, within two months: older age (aHR 1.3; P = 0.001) and a previous ankle injury (aHR 3.9; P = 0.032). For a reinjury within two seasons: age (aHR 1.1; P = 0.013) and above all a previous calf injury, with an adjusted hazard ratio of 6.7 (P = 0,002) (Green 2020, PMID 33030961). Consistency with the European registry is good: across 22,942 injuries in the UEFA cohort, calf injury is among the six diagnoses for which reinjury significantly lengthens the absence compared with the index injury (Ekstrand 2020, PMID 31182429).

How do you manage loading in the first few days?

Once the two emergencies are ruled out, the question becomes ordinary and is handled with the principles common to any muscle injury. This chapter therefore stays deliberately short: the pathophysiology of muscle healing and the general principles of return to loading are covered elsewhere on the site.

For the biology of muscle repair, the inflammatory, repair and remodelling phases, and the general logic of progressive reloading after injury, see muscle contusion (dead leg) and its complications, which covers the direct-blow side, and proximal hamstring tendinopathy for the progressive loading side of a thigh muscle-tendon complex. What follows covers only what is specific to the triceps surae.

The first three days

The calf's specificity at this stage lies in a simple mechanical constraint: it is the muscle you cannot unload without unloading walking. Unlike a hamstring or an adductor, the triceps surae works at every step, eccentrically at heel strike and concentrically at push-off. Loading is therefore not an exercise to be added to the programme: it is the patient's basic activity, and it is that which has to be dosed.

The most useful time landmark comes from the Australian registry: the median to walk without pain after a medial gastrocnemius injury is 3 days (interquartile range 3 days) (Gray 2025, PMID 40985375). A patient still unable to walk normally at the end of the first week is off the usual trajectory: that is a reason to re-examine, not to wait.

What rehabilitation for calf injury, and at what level of evidence?

A warning at the outset: the treatment literature specific to the calf is thin. There is no randomised controlled trial of triceps surae injury rehabilitation comparable with those that exist for the hamstrings. This chapter says so rather than hiding it behind firm recommendations.

That gap is not an opinion: it is the observation that opens the qualitative study of 20 international expert clinicians, which begins by noting “the scarcity of research to guide clinicians facing these injuries” (Green 2022, PMID 35032233). That study is, paradoxically, the best source of practice available, but it is structured expert opinion, not experimental evidence.

What the experts do, and what guides them

Analysis of the interviews brings out several converging points. Calf injuries are considered different from other muscle injuries and treated as such. Rigorous clinical work establishes the diagnosis; what gives the most reliable prognosis, on the other hand, is not the initial imaging but monitoring of calf capacity and its response to load exposure. Rehabilitation is structured in six phases, guided by principles aimed at optimising performance on return to play while avoiding reinjury.

Six-phase structure after Green B, McClelland JA, Semciw AI, et al. Sports Med Open 2022;8(1):10 (PMID 35032233), qualitative study of 20 expert clinicians. The quantified landmarks are those of Gray C, et al. J Orthop Sports Phys Ther 2025;55(10):681-8 (PMID 40985375). The exit criteria proposed are a practical arrangement, not validated as such.

The modalities, one by one, with their level of evidence

The table below applies GRADE logic to the modalities used on the calf. The reader will note that no line reaches the “high” level. That is a faithful reflection of the state of the literature on this site, and not stylistic caution.

How do you strengthen the triceps surae without forgetting the soleus?

This is the most operational chapter of the article, and the one where the commonest error is also the easiest to correct: strengthening the calf only with the knee extended means leaving out the muscle that carries most injuries and nearly all reinjuries.

Why the flexed knee is not a variant but a necessity

The argument is mechanical and it is measurable. When the knee flexes, the gastrocnemius, biarticular, loses length and therefore force-producing capacity. The soleus, monoarticular, is unaffected. The result is that knee flexion shifts the load of the complex towards the soleus.

That shift has been quantified. In nine physically active men, maximal isometric plantarflexion torque went from 154.4 ± 37.8 N·m with the knee extended to 93.1 ± 22.3 N·m with the knee flexed. And above all, the correlation analysis shows that knee position changes the relationship between muscle size and torque produced for the gastrocnemius but not for the soleus : only soleus volume and total triceps surae volume remained correlated with torque produced with the knee flexed (Kovács 2024, PMID 39577118).

The reservation about the sample deserves stating: nine subjects, all active men. The absolute values are not meant to serve as an individual reference. The direction of the effect, on the other hand, is consistent with the anatomy and with the whole literature on triceps surae architecture, and it is that direction that governs practice.

The progression, exercise by exercise

The logic is constant: tolerance first, then volume, then load, then speed. Each stage is done in both knee positions, and you progress in the two independently: it is common for a patient to tolerate load with the knee extended well before tolerating it with the knee flexed, or the reverse depending on the muscle involved.

On what criteria should return to sport be allowed?

The question is asked back to front in most consultations: “how long?” rather than “on what conditions?”. The reinjury data give a quantified reason to reverse the question.

The most useful reminder is this: in a cohort of 563 calf injuries followed over ten years, 20 % of subsequent injuries occurred before full recovery, and those relapses cost 46.7 ± 25.6 days to return, against about twenty days for an index injury (Green 2025, PMID 40919406). Returning a week too early does not gain a week: it costs three.

The state-of-the-art review on tennis leg puts the same thing in terms of principle: recovery usually occurs in 6 to 12 weeks, and return to activity must be guided by functional recovery, not by the disappearance of symptoms alone (Schwach 2026, PMID 42258915).

The reference test: single-leg heel raises

The heel rise endurance test is the best tool the physiotherapist has for objectifying calf capacity, because it is reproducible, needs no equipment, and has just been given normative values in a large population.

An international study of 500 subjects with no history of Achilles tendinopathy or recent immobilisation gives medians of 25 repetitions on the dominant side and 24 on the non-dominant side, for total work of 1,374 and 1,325 J, cumulative vertical displacement of 192 and 186 cm, and a peak height of 9.3 and 9.6 cm. The authors note lower performance with female sex, higher BMI and lower physical activity level (Visser 2025, PMID 40020545).

The most operational result of that study is not the median of 25 repetitions, it is the absence of a significant difference between dominant and non-dominant side. It validates the practice of side-to-side comparison: in a patient with a unilaterally injured calf, the sound side is an acceptable reference, which avoids having to correct for sex, BMI and activity level.

The criteria grid

What do concrete clinical cases teach us?

Three published cases, chosen because each illustrates a different error and because all three are genuine referenced case reports. None is reconstructed or composite.

Case 1: When imaging concludes to a thrombosis and gets it wrong

A 51-year-old man, playing tennis, feels as though he has been kicked in the back of the calf, with an audible snap and severe pain. Ultrasound finds fluid between the muscle planes and a hypoechoic structure; the radiology registrar concludes to a deep vein thrombosis. MRI corrects the diagnosis: rupture of the plantaris, with the fluid and the muscle stump visible between gastrocnemius and soleus. The patient is treated conservatively (rest, ice, compression, elevation), with clear resolution of the pain and the swelling (Rohilla S, Jain N, Yadav R. BMJ Case Rep 2013;2013:bcr-2012-007840, PMID 23345486).

What this case teaches. The error goes the other way too: here it is a muscle injury that was taken for a thrombosis. The authors note in passing a second trap in the same picture: an intact plantaris can still produce plantarflexion when the Achilles tendon is ruptured, which gives “a misleading picture”. The practical corollary is that of chapter 3: never conclude that the tendon is intact from the persistence of active movement alone.

Case 2: The ordinary form, and its real trajectory

A 35-year-old man, sudden intense calf pain after a minor injury with the leg extended. The diagnosis of tennis leg, with a partial gastrocnemius tear, is made on clinical examination and ultrasound. The authors explicitly stress that ultrasound served to distinguish it from a deep vein thrombosis. Conservative management: rest, ice, compression, elevation, analgesia, physiotherapy. Normal muscle function at the four-month review (Adhikari S, Devkota S, Lamichhane S, et al. SAGE Open Med Case Rep 2024;12:2050313X241272690, PMID 39430721).

What this case teaches. Two things, one reassuring and one demanding. The reassuring one: the prognosis of conservatively treated tennis leg is good, and surgery is reserved for severe forms only. The demanding one: follow-up ran to four months. The time to return to sport published in the cohorts, about twenty days in the elite athlete supervised daily, does not describe the trajectory of the clinic patient. The injury mechanism was moreover “minor”, which is a reminder that the intensity of the trauma does not predict that of the injury.

Case 3: Both diagnoses at once

A patient with an established and obvious tennis leg subsequently develops thrombophlebitis in the same limb. The author, who reports this case as the first of its kind published, recalls in the same article that the earlier literature documented mainly the reverse error: tennis legs taken for thrombophlebitis, treated with unnecessary anticoagulation, with bleeding complications as a result (Slawski DP. J Orthop Trauma 1994;8(3):263-4, PMID 8027900).

What this case teaches. It is the most important of the three, and the one that defeats the most natural reasoning. Having established a muscle injury does not protect against a thrombosis: the two can coexist, and the relative immobilisation that follows the injury is itself a thrombogenic factor. Screening for red flags is not done once, at the first session; it is done again at every session, particularly if the swelling increases again or if the pain changes in character.

How do you apply all this concretely in practice?

The operational summary, in the order of the consultation.

At the first session

1. Reconstruct the moment of injury. A triceps surae injury nearly always has its movement and its second. Their absence is the first red flag, before any physical examination.
2. Go through the red flag grid from chapter 2: thrombogenic background, the time course of the swelling, night pain, systemic signs. A single feature is enough to suspend local treatment.
3. Test the Achilles tendon, routinely and with at least two tests : calf squeeze and Matles, plus comparison of the resting equinus. In every ruptured patient in Maffulli's series, at least two tests were positive.
4. Locate the muscle by resisted contraction with the knee extended then flexed, and by level-by-level palpation of the medial myotendinous junction.
5. Measure before treating : bilateral calf circumference, dorsiflexion in both knee positions, number of heel raises on the sound side. Those three measurements become the exit criteria, and it will be too late to take them afterwards.
6. Announce an order of magnitude, not a date, and explain why: the published intervals are wide and the cost of an early return is higher than that of a week's wait.

What you do not do

• Do not massage or compress a calf in which thrombosis has not been ruled out. One patient in ten referred for tennis leg had an isolated DVT.
• Do not conclude that the Achilles tendon is intact from palpation alone (sensitivity 0.73), nor from the persistence of active plantarflexion.
• Do not strengthen only with the knee extended. That leaves out the soleus, which carries most injuries and reinjuries.
• Do not take painless walking on the third day for recovery. It is a documented median, not a return criterion.
• Do not stop monitoring the red flags once the diagnosis has been made: thrombosis can occur on top of a genuine muscle injury.

What you write in the notes

Frequently asked questions

“Tennis leg” and calf injury, are they the same thing?

Not exactly. “Tennis leg” classically denotes injury of the medial head of gastrocnemius at its myo-aponeurotic junction. It is one form of calf injury, but not the only one: in the field athlete, the soleus is at fault in 84.6 % of cases (Green 2020, PMID 31494970). Using “tennis leg” as a synonym for “calf injury” leads to neglecting the muscle most often involved.

Is tennis leg a rupture of the plantaris?

That idea, long taught, does not stand up to the data. Across 141 patients referred with a clinical diagnosis of tennis leg, the plantaris was at fault in only 2 cases, that is 1.4 %; the medial head of gastrocnemius was in 66.7 % (Delgado 2002, PMID 12091669). In the same series, an isolated deep vein thrombosis was seven times commoner than plantaris involvement.

How can you be sure it is not a clot?

You cannot be sure clinically, and that is precisely the message of this chapter. No examination sign rules out a deep vein thrombosis, and the Wells score itself is not enough: in a cohort of 1,135 inpatients, 5.9 % of those classed as low probability had a proximal DVT (Silveira 2015, PMID 25985219). The physiotherapist's role is to spot the doubt, suspend all local treatment and refer for duplex ultrasound, not to decide.

How long before returning to sport?

The cohort landmarks, in elite athletes supervised daily, give medians of 19 days for a gastrocnemius injury (Gray 2025, PMID 40985375) and a mean of 25.4 ± 16.2 days for the soleus (Green 2020, PMID 31494970). The reviews on tennis leg suggest rather 6 to 12 weeks in the general population (Schwach 2026, PMID 42258915). These figures diverge because they describe different populations and different return criteria. That is one more reason to decide on functional criteria and not on a calendar.

Why does return depend so much on the connective tissue?

Because it is what transmits the muscle's force to the tendon. Across 100 consecutive patients, the mean time to return ran from 8 days when the connective interfaces were intact to 48 days when they were disrupted, that is a sixfold difference within the same condition (Prakash 2018, PMID 29074478). The extent of muscle oedema, for its part, is a poorer predictor.

Should an ultrasound be done routinely?

No, but it has two advantages that often make it useful first line: it locates the injury and, above all, it sees deep vein thrombosis, which settles the safety question at the same time. The great majority of calf injuries can be diagnosed and managed with no further imaging (Bright 2017, PMID 28661826). MRI is reserved for situations where a precise prognosis has a consequence, or where the course does not follow.

Why work with the knee flexed?

Because it is the only position that loads the soleus selectively. The gastrocnemius crosses the knee: when it flexes, the muscle shortens and loses force-producing capacity. Maximal isometric plantarflexion torque goes from 154.4 to 93.1 N·m between knee extended and knee flexed, and the relationship between muscle volume and torque produced stays unchanged for the soleus while it changes for the gastrocnemius (Kovács 2024, PMID 39577118).

My patient has already had an injury in the same calf. Should I do things differently?

Yes, in how demanding the exit criteria are rather than in the content of the exercises. A previous calf injury multiplies the reinjury risk over the following two seasons by 6.7 (Green 2020, PMID 33030961), and along with age it is one of only two solidly established risk factors (Green & Pizzari 2017, PMID 28259848). In those patients, you require measured full symmetry, not merely the absence of complaint.

Can a muscle injury and a thrombosis occur together?

Yes, and it is documented: a case of ipsilateral thrombophlebitis occurring on an already established tennis leg has been reported (Slawski 1994, PMID 8027900). That is why red flag screening is repeated at every session, particularly if the swelling increases again after a phase of resolution.

Does a patient who can still move the foot necessarily have an intact Achilles tendon?

No. The toe flexors, tibialis posterior, the peroneals and the plantaris still produce plantarflexion when the Achilles tendon is ruptured. It is a classic trap, flagged in the case literature (Rohilla 2013, PMID 23345486). Only the specific tests decide, and at least two are needed.

The sources for this article

Every reference was verified on PubMed, and the content of each abstract read to ensure that it does establish what is attributed to it in the text. The identifiers are clickable.

1. Delgado GJ, Chung CB, Lektrakul N, et al. Tennis leg: clinical US study of 141 patients and anatomic investigation of four cadavers with MR imaging and US. Radiology 2002;224(1):112-9. PMID 12091669
2. Green B, Pizzari T. Calf muscle strain injuries in sport: a systematic review of risk factors for injury. Br J Sports Med 2017;51(16):1189-94. PMID 28259848
3. Green B, Lin M, Schache AG, et al. Calf muscle strain injuries in elite Australian Football players: a descriptive epidemiological evaluation. Scand J Med Sci Sports 2020;30(1):174-84. PMID 31494970
4. Green B, Lin M, McClelland JA, et al. Return to play and recurrence after calf muscle strain injuries in elite Australian Football players. Am J Sports Med 2020;48(13):3306-15. PMID 33030961
5. Green B, McClelland JA, Semciw AI, et al. The assessment, management and prevention of calf muscle strain injuries: a qualitative study of the practices and perspectives of 20 expert sports clinicians. Sports Med Open 2022;8(1):10. PMID 35032233
6. Green B, Schache AG, Pizzari T. What is a recurrence? The onset, frequency and time loss impact of recurrent calf muscle strain injuries in elite male Australian football players over a decade. BMJ Open Sport Exerc Med 2025;11(3):e002865. PMID 40919406
7. Gray C, Pizzari T, Murphy MC, et al. Gastrocnemius muscle strain injury characteristics in elite male Australian football players: a 10-year longitudinal cohort study. J Orthop Sports Phys Ther 2025;55(10):681-8. PMID 40985375
8. Prakash A, Entwisle T, Schneider M, et al. Connective tissue injury in calf muscle tears and return to play: MRI correlation. Br J Sports Med 2018;52(14):929-33. PMID 29074478
9. Pedret C, Balius R, Blasi M, et al. Ultrasound classification of medial gastrocnemious injuries. Scand J Med Sci Sports 2020;30(12):2456-65. PMID 32854168
10. Pezzotta G, Querques G, Pecorelli A, et al. MRI detection of soleus muscle injuries in professional football players. Skeletal Radiol 2017;46(11):1513-20. PMID 28770309
11. Mueller-Wohlfahrt HW, Haensel L, Mithoefer K, et al. Terminology and classification of muscle injuries in sport: the Munich consensus statement. Br J Sports Med 2013;47(6):342-50. PMID 23080315
12. Bright JM, Fields KB, Draper R. Ultrasound diagnosis of calf injuries. Sports Health 2017;9(4):352-5. PMID 28661826
13. Schwach M, Gaulin B, Vermorel PH, et al. Tennis leg: diagnosis and management: a state-of-the-art review. Foot (Edinb) 2026;68:102267. PMID 42258915
14. Fields KB, Rigby MD. Muscular calf injuries in runners. Curr Sports Med Rep 2016;15(5):320-4. PMID 27618240
15. Maffulli N. The clinical diagnosis of subcutaneous tear of the Achilles tendon. A prospective study in 174 patients. Am J Sports Med 1998;26(2):266-70. PMID 9548122
16. Garras DN, Raikin SM, Bhat SB, et al. MRI is unnecessary for diagnosing acute Achilles tendon ruptures: clinical diagnostic criteria. Clin Orthop Relat Res 2012;470(8):2268-73. PMID 22538958
17. Kotsifaki R, Malliaras P, Byron C, et al. Incidence, temporal trends, and surgical shift of Achilles tendon rupture: a systematic review and meta-analysis. Sports Med 2026;56(6):1467-87. PMID 41933260
18. Ekstrand J, Krutsch W, Spreco A, et al. Time before return to play for the most common injuries in professional football: a 16-year follow-up of the UEFA Elite Club Injury Study. Br J Sports Med 2020;54(7):421-6. PMID 31182429
19. Silveira PC, Ip IK, Goldhaber SZ, et al. Performance of Wells score for deep vein thrombosis in the inpatient setting. JAMA Intern Med 2015;175(7):1112-7. PMID 25985219
20. Righini M, Robert-Ebadi H. Management of isolated distal deep vein thrombosis. Vasa 2024;53(3):185-92. PMID 38546285
21. Sato O, Kondoh K, Iyori K, et al. Midcalf ultrasonography for the diagnosis of ruptured Baker's cysts. Surg Today 2001;31(5):410-3. PMID 11381504
22. Slawski DP. Deep venous thrombosis complicating rupture of the medial head of the gastrocnemius muscle. J Orthop Trauma 1994;8(3):263-4. PMID 8027900
23. Rohilla S, Jain N, Yadav R. Plantaris rupture: why is it important? BMJ Case Rep 2013;2013:bcr-2012-007840. PMID 23345486
24. Adhikari S, Devkota S, Lamichhane S, et al. Trivial incident leading to tennis leg managed conservatively: a case report. SAGE Open Med Case Rep 2024;12:2050313X241272690. PMID 39430721
25. Visser TSS, Neill SO, Hébert-Losier K, et al. Normative values for calf muscle strength-endurance in the general population assessed with the Calf Raise Application: a large international cross-sectional study. Braz J Phys Ther 2025;29(3):101188. PMID 40020545
26. Kovács B, Yaodong G, Kóbor I, et al. Effects of knee joint position on the triceps suræ torque-size relationship during plantarflexion in healthy young adults. J Biomech 2024;177:112436. PMID 39577118

On the same clinical axis

• Achilles tendinopathy: chronic distress of the same muscle-tendon complex, and the differential of non-acute posterior pain.
• Muscle contusion (dead leg) and its complications: muscle injury from a direct blow, its pathophysiology and its own complications.
• Proximal hamstring tendinopathy: progressive loading of a lower limb muscle-tendon complex.
• Baker's cyst (popliteal cyst): the differential of pseudo-thrombophlebitis.
• Medial tibial stress syndrome (MTSS) in the runner: leg pain of gradual onset on exertion.
• Lateral ankle sprain: a previous ankle injury multiplied the risk of early calf reinjury by 3.9.