Training guide
There's no single correct way to run. Beginners become more economical on their own within weeks, form explains only a small share of why some runners use less energy than others, and most runners land on their heel without trouble. The changes with the best evidence are small ones: a slightly quicker step (5 to 10% more steps a minute) and landing with your foot closer under you, which both reduce the load on the knee in lab studies.
This guide covers what good form means, cadence and the myth of 180 steps a minute, overstriding, foot strike, posture and arm swing, what gait retraining has and hasn't shown for injuries, drills and strides, and how to change one thing at a time without swapping one pain for another.
Ten beginners became about 8.5% more economical in 10 weeks of running with no coaching on form, and across 51 studies, single parts of form explained only a small share of the differences in economy between runners. If running feels fine and you're not injured, you don't need to change anything.
Taking 5 to 10% more steps a minute at the same speed reduced the load on the knee in lab studies, mostly by shortening the stride so the foot lands closer under you. Whether it prevents injuries hasn't been shown.
Runners' own step rates vary widely, from about 155 to over 200 a minute even among elite runners. Step rate rises with speed and falls with height and leg length.
About 9 in 10 runners in a road race landed heel first, and no study that followed runners forward has linked foot strike with injury. Switching to a forefoot strike made running less economical at first and moves load to the calf and Achilles.
In one trial of 320 novice runners, 2 weeks of feedback to land more softly went with fewer self-reported injuries over a year (16% against 38%). Across 19 trials, too few measured injury to say more.
No trial has shown that running drills improve form or prevent injury. Strength training does make running 2 to 8% more economical in most studies. If you change something, change one thing at a time over 2 to 4 weeks, and stop for pain.
Form is how you move when you run: your step rate (cadence, steps a minute), stride length, where your foot lands compared with your hips, how much you bounce, how upright your trunk is and how your arms swing. Speed is step rate times step length, so at the same speed a quicker step is a shorter one.
Form matters to runners in two ways: economy (the energy a given pace costs) and the load on particular tissues, such as the knee, the kneecap, the Achilles and calf, and the shin. A change often moves load from one place to another rather than removing it: a forefoot landing eases the knee but loads the calf and Achilles.
The calf muscles do much of the supporting work at running speeds, the thigh muscles absorb the landing, and the hip muscles steady the pelvis on each one-leg landing, as our running page explains. Injured runners showed more drop of the opposite hip on landing than uninjured runners.
Source 19
Height, leg length and speed change step rate; beginners differ from trained runners in how far their chosen stride is from their most economical one; most runners land heel first; and shoes and surfaces change how you land. Being able to jog easily for 20 to 30 minutes comes before drills and strides, and pain-free running comes before any change aimed at performance.
Ask someone to film you from the side and from behind on an easy run, on flat ground. Count your steps for 30 seconds and double it to get your step rate. Look for a foot that lands well ahead of your hips, a big up-and-down bounce, arms swinging across your body, and a hip that drops on the other side each time you land (our suggestion).
If nothing hurts and running feels fine, nothing needs fixing. If you see a long reach with your foot, try the slightly quicker step below. If you have pain, start with the matching pain guide or a physiotherapist rather than changing your form on your own (our suggestion).
In 10 women who were new to running, 10 weeks of running with no coaching on form cut the oxygen they used per kilometre by about 8.5% (our arithmetic), while their maximum oxygen use didn't change. Several parts of their stride changed on their own.
Source 1
Across 51 studies of 1,115 people, single parts of running form each explained only about 4 to 12% of the differences in economy between runners, and a higher step rate had only a small link with lower energy cost. In 97 runners, three technique measures together explained 39% of the energy cost. A review warns that "recommending a general economical running technique should be approached with caution".
In trained runners, the stride they choose is about their most economical, and strides up to 3% shorter are fine. In novices, the chosen and most economical step rates can differ more, by about 8%.
Source 3
Across 24 studies of middle- and long-distance runners, strength training generally improved running economy by 2 to 8% in the 20 that measured it, though not in every study. Heavy lifting, jumping (plyometric) training and the two combined did most.
In 45 recreational runners, 5% more steps a minute at the same speed cut the energy absorbed at the knee by about 20%, and 10% more cut it by about 34% and eased the hip too. In 30 runners, 10% more steps cut the modelled peak force on the kneecap by 14%. Across 37 studies, the evidence that a quicker step means less knee bend was strong, and for less kneecap stress, limited. These are lab findings about load, not about injuries.
The same review found "insufficient evidence to conclusively determine the effects of altering running step rate on injury and performance", and very limited evidence that a quicker step feels more effortful and awkward and costs a little more energy.
Source 6
In 30 recreational runners, own step rates ran from 156 to 192 a minute. Outdoors, 21 experienced runners averaged 165 at a comfortable speed, a little below the 171 that kept their heart rate lowest, and both rose by 3 steps a minute when they ran 15% faster; taller runners took fewer steps. Elite runners in a 100 km race averaged 182, ranging from 155 to 203. In 138 young runners, leg length and speed explained about half the variation.
Overstriding means landing with your foot far ahead of your hips. In 14 runners, braking force rose the further ahead of the hip the foot landed, and for every 5 more strides a minute, the foot landed about 6% closer to the hip. A quicker step shortens the stride and brings the foot back under you.
Of 936 runners in a road race, mostly recreational, about 89% were landing heel first at the 10 km point.
Source 9
Across 13 studies, the links between foot strike and injury came from studies that looked back after the injuries, which can't show cause; the few that followed runners forward found no link. In one of the looking-back studies, of 52 college runners, heel strikers reported about twice the rate of overuse injuries, but the authors say it doesn't test cause.
When heel strikers were asked to land on their forefoot, running became less economical straight away at slow and moderate speeds, and the load shifted to the ankle and calf. A review concluded that "changing strike pattern cannot be recommended for an uninjured RFS runner" (RFS: a rearfoot, or heel, striker).
In 320 novice runners at a running club, 2 weeks of treadmill running with real-time feedback to land more softly went with 16% injured over the next year, against 38% in runners who did the same treadmill time without feedback. It's one trial, the injuries were self-reported, and it hasn't been repeated, so it doesn't show that other changes to form prevent injury.
Source 12
Across 19 trials of 673 people, step-rate retraining raised step rate and lowered how quickly force builds up on landing (moderate certainty), without changing economy or performance. Too few trials measured pain; two showed fewer injuries at a year.
Source 13
In 69 runners with kneecap pain, education on symptoms and training load did as well on its own as with added gait retraining or exercises, though the retraining did raise step rate by 7% and cut the loading rate by 25%. A review of reviews found short-term gains in pain and function from 5 to 10% more steps or a softer landing, and in 12 runners with kneecap pain, a 10% quicker step went with less pain at 3 months, without a comparison group.
Programmes used about 8 sessions over 2 weeks, building from 15 to 30 minutes of running, with the feedback faded over the last 4 sessions; in 10 runners, this cut impact at the shin by nearly half a month later. Outdoors, 30 runners cued to a 7.5% quicker step over 8 sessions raised it by 8.6% and cut the loading rate by about 19% at 1 month. Most changes were still there 1 to 3 months later.
In 72 injured and 36 uninjured runners, the injured showed more drop of the opposite hip, more forward trunk lean and a straighter knee on landing, whichever of 4 common injuries they had. This compares runners at one point in time, so it can't show which came first.
Source 19
In 24 runners, a little more forward lean of the trunk (about 14 degrees against their usual 7) lowered peak stress on the kneecap, and leaning back raised it. In 97 runners, a more upright trunk went with better performance. Our reading: stay tall and let a slight lean come from the ankles, not by bending at the waist.
Running with the arms held still behind the back, across the chest or on the head cost 3%, 9% and 13% more energy than running with a normal arm swing. Keep the elbows bent to about a right angle, hands and shoulders relaxed, swinging forward and back rather than across your body (our suggestion).
Source 34
In 17 long-distance runners, A-skips, B-skips and heel flicks bent the ankle less than running did. In 30 recreational runners, adding drills to 15 weeks of interval training didn't improve performance, and one loading measure was higher. No trial has shown that drills improve form or prevent injury: they're a way to practise, and our drills session is untested.
Across 12 trials of 11,240 people, no type of running shoe reduced injuries, and choosing shoes by foot posture probably made little or no difference (moderate certainty). In 848 runners randomly given softer or harder shoes, the harder ones went with more injuries, but only in lighter runners when split by weight. Choosing a comfortable shoe is a sensible idea, not a tested one.
Our editorial sequence. The 5 to 10% steps come from the step-rate studies (Heiderscheit 2011; Lenhart 2014) and the retraining studies (Bramah 2019, 10% in runners with kneecap pain; Willy 2016; the Anderson 2022 meta-analysis); the 2 to 4 weeks and fading the feedback follow the gait retraining studies above (Crowell 2011; Willy 2012). It hasn't been tested as a programme.
Loading the 3D figure…
The movement on the figure: Treadmill running
The retraining studies changed one thing over about 8 sessions in 2 weeks and faded the feedback towards the end. Give one change 2 to 4 weeks on easy runs before you judge it or add another, and keep your weekly distance the same while you do (our suggestion).
A quicker step felt more awkward and effortful in a few small studies (very limited evidence). Keep the change on easy runs until it stops feeling strange, and leave speed sessions as they are (our suggestion).
Source 6
If a new aching starts in your calf, Achilles, shin or foot after you change how you land or step, go back to your old running for a week or two and build the change in more slowly. Pain that doesn't settle needs its pain guide (our suggestion).
After an easy jog, once or twice a week; doses from our runner's warm-up and speed workouts. Count each landing in the hops and skips as a foot contact: about 80 to 100 a session suits beginners, as our fast skipping page says (Davies 2015).
Doses from our injury prevention exercises for runners, done 3 or 4 times a week (single-leg calf raises 3 × 15 each leg, single-leg deadlifts 2 × 10 each leg, side plank leg lifts 2 × 10 each side, single-leg balance for 2 minutes each leg), and our strength workout for runners at home, done twice a week (split squats 8 to 12 each side).
3D demoSingle-leg calf raise from a stepIntermediate · A step + A wall
3D demoSingle-leg Romanian deadliftAdvanced · A dumbbell
3D demoSide plank with leg liftAdvanced · No equipment
3D demoRear-foot-elevated split squat (Bulgarian split squat)Intermediate · A bench or box
3D demoSingle-leg standBeginner · A wallBefore runs, or daily if your ankles or hips feel stiff: ankle circles and knee-to-wall rocks as their move pages give them (5 slow circles each way with each ankle; 10 slow rocks per side, daily), and the hip flexor stretch for 45 seconds each side after a run, as in our post-run stretches.
Our running form drills session takes 25 minutes: an easy jog, A-skips, fast skips, an arm drill, small hops and strides, then an easy run with one cue. Do it once or twice a week, after an easy run or in place of one. It's our editorial session and hasn't been tested.
A 25-minute session to practise running tall, quick and relaxed: an easy jog, A-skips, fast skips, an arm drill, small hops, strides and an easy run with one cue.
25 minBeginner10 movesCardio machineAn 8-minute dynamic warm-up for runners: ankle and hip mobility, lunges, skips and pogo hops.
8 minIntermediate8 movesA wallA 20-minute hop and bound session for runners: ankle bounces to pogo hops, one-leg hops and bounding, building foot contacts slowly over 10 weeks.
20 minIntermediate8 movesCardio machineA 40-minute strength workout for runners at home: hops, Bulgarian split squats, single-leg deadlifts, step-ups, calf raises and side planks.
40 minIntermediate15 movesA wall, A bench or boxA 45-minute gym strength workout for runners: hops, heavy squats, Bulgarian split squats, calf raises and side planks, with sets, reps and rest.
45 minIntermediate13 movesA wall, A bench or boxFoot, calf, balance and hop exercises taken from the running trials with positive results, in a 25-minute home routine for 3 or 4 days a week.
25 minBeginner11 movesOther / check instructions, A bench or boxA 45-minute tempo run for 10K and half marathon training: an easy warm-up and strides, then two 8-minute blocks at a comfortably hard, short-sentences effort.
45 minIntermediate6 movesCardio machine, A wallThe strength training plan for runners builds the strength work over weeks; the 5K plan adds strides and hills to easy running.
A 12-week strength plan for runners: two sessions a week of heavy lifts, single-leg work, calf raises and hops, fitted around your runs.
2 training days45 minBeginner4 days a weekAn 8-week plan to run a faster 5K once you can run one: hills, then intervals and tempo runs, mostly easy running, strength, a time trial and a light race week.
4 training days40 minIntermediateWith kneecap pain, start with the runner's knee or knee pain when running guide: advice on symptoms and training load came first in a trial, and a 5 to 10% quicker step or softer landing is added if needed, often with a physiotherapist.
For exertional compartment syndrome at the front of the shin, changing how you run is the core of treatment without surgery, as our running page explains. Experts also use running retraining for several other leg injuries, despite little trial evidence. That's a plan to make with a physiotherapist and the pain guide, not on your own.
Source 38
Our shin splints, Achilles tendinopathy, plantar fasciitis and IT band syndrome guides each have a staged plan. Load comes first in each; form changes, if any, come later.
If you get chest pain or pressure, feel faint or dizzy, are unusually breathless, or your heart races or beats irregularly during exercise, as our plans advise.
For chest pain that comes on suddenly and doesn't go away, feels tight or like squeezing, spreads to your arms, neck, jaw, stomach or back, or comes with sweating, feeling sick, light-headedness or breathlessness; severe difficulty breathing; someone who faints while exercising; palpitations that don't go away, or come with chest pain, shortness of breath or feeling faint; or a headache that starts suddenly and is extremely painful, especially while lifting or straining, or comes with numbness or weakness, difficulty speaking or loss of vision. The NHS says a bleed on the brain may be more likely when lifting something heavy. Don't drive yourself to A&E.
Stiff, achy legs for a day or two are normal at first and ease as you repeat the runs. Stop and rest if a pain is sharp, makes you limp or change how you run, or builds as you go, and skip your next run if it's worse the next morning. The NHS advises not running when you're in pain, and seeing a GP or physiotherapist if it doesn't go away; a very painful or swollen knee, or a sudden sharp pain at the back of the ankle, needs a GP straight away.
Source 41
Tightness in the shin or calf, with tingling on the top of the foot or a foot that slaps down, that comes on at the same point in every run and eases within minutes of stopping: see a GP or sports physiotherapist. It can be exertional compartment syndrome.
Pain in one spot on the shin or a foot bone that you can point to with a fingertip, worse with impact and later present at rest or at night: a possible stress fracture. Stop running and jumping until it has been checked.
Call 999 or go to A&E: loss of feeling or numbness around the genitals or bottom, new trouble peeing or controlling your bladder or bowels, changes in genital sensation during sex or problems with erections or orgasm, or pain, tingling, weakness or numbness in both legs. These are signs of cauda equina syndrome, which needs treating in hospital urgently. Don't drive yourself.
Pain, tingling or numbness spreading below the knee that isn't settling after a few weeks, is getting worse or stops your normal activities: see a GP.
If a new aching starts in your calf, Achilles, shin or foot after you change how you land or step, go back to your old running for a week or two and build the change in more slowly. Pain that doesn't settle needs its pain guide (our suggestion).
Ask for an urgent GP appointment or get help from NHS 111 for a headache that's triggered or made worse by coughing, sneezing, bending down or exercising. In the US, contact your doctor or urgent care.
Speak to a GP first if you haven't exercised for some time, or if you have a medical condition or concerns. Get advice first, too, if you have a health problem, an injury or symptoms, have recently had a heart attack or an operation, or are pregnant or have recently given birth.
Answer the 7 questions of the Physical Activity Readiness Questionnaire (PAR-Q+) before you start. A yes to any (a heart condition or high blood pressure, chest pain, dizziness or fainting, another long-term condition or its medicines, a bone or joint problem, or being told to exercise only under medical supervision) means checking with your doctor or a qualified exercise professional first. It also asks people over 45 who aren't used to hard, all-out effort to check before exercising that hard: think races and time trials, not easy runs. The NHS suggests speaking to a GP first if you're worried about your health.
Source 42
If you have a heart condition, follow your cardiac team's advice on how hard to work, and stop for the warning signs above. People with long-term conditions may wish to ask their specialist before starting. With uncontrolled high blood pressure, avoid straining against a held breath; the high blood pressure guide has more.
About 9 in 10 road racers land heel first, and no study that followed runners forward has linked foot strike with injury. The links come from studies that looked back after injuries, which can't show cause.
It made running less economical at first and moves load to the calf and Achilles, so it isn't recommended if you're not injured.
Recreational runners' own rates ran from 156 to 192, and step rate rises with speed and falls with height and leg length. Elite ultrarunners ranged from 155 to 203.
One trial of soft-landing retraining in new runners found fewer self-reported injuries; otherwise, the evidence on form and injury is thin.
Drills don't copy running mechanics, and adding them to training didn't improve performance in the one trial.
Beginners become more economical on their own, and single parts of form explain only a small share of the differences in economy between runners.
No type of running shoe has been shown to prevent injuries, and choosing shoes by foot posture made little or no difference.
Source 35
A little more lean lowered kneecap stress in one lab study, while a more upright trunk went with better performance in another. Stay tall and let a slight lean come from the ankles.
There's no single correct form. In general: run tall with a slight lean from the ankles, land with your foot near your hips rather than far in front, take quick, light steps, and let your arms swing relaxed. Beginners become more economical on their own within weeks.
Run regularly at an easy pace, strength train twice a week, and if you want to change one thing, try a step rate about 5% quicker at the same speed. Change one thing at a time over 2 to 4 weeks, and stop if it brings on pain.
Whatever is natural for your height and speed: recreational runners' own rates ran from about 156 to 192 steps a minute. A rate 5 to 10% above your own reduced the load on the knee in lab studies, though it hasn't been shown to prevent injuries.
No. Cadence rises with speed and falls with height and leg length, and even elite ultrarunners ranged from 155 to 203 steps a minute in a race.
No. About 9 in 10 runners in a road race landed heel first, and no study that followed runners forward has linked heel striking with injury. Overstriding, landing with your foot far ahead of your hips, is the part worth changing.
Not if you're not injured: switching made running less economical at first and moves load to the calf and Achilles. It's a treatment option for some injuries, planned with a physiotherapist.
It's landing with your foot well ahead of your hips, which increases braking. A slightly quicker step at the same speed shortens your stride and brings your foot back under you.
Not proven. A quicker step and a softer landing reduce knee load and helped knee pain in small studies, but advice on training load did as well in a trial. In one trial in new runners, landing more softly went with fewer self-reported injuries. Our pain guides have the full plans.
Let them swing forward and back, elbows bent to about a right angle, hands and shoulders relaxed. Holding the arms still cost 3 to 13% more energy in a lab study.
No trial has shown that they do. Drills don't copy running mechanics, and adding them to training didn't improve performance in the one trial. They're useful practice; strength training improves running economy more reliably.
Retraining studies used about 8 sessions over 2 weeks, and most changes were still there 1 to 3 months later. Allow 2 to 4 weeks for each change on easy runs before you add another.
No type of running shoe has been shown to prevent injuries, and choosing shoes by foot posture made little or no difference. Pick a pair that feels comfortable.
Exercise education, not a personal prescription, and clinical review of these guides is pending. Check with a GP before starting hard training if you have a heart condition, chest pain, dizziness or breathlessness when active, and stop for sharp or increasing pain. Health disclaimer.
