Strength Training for Runners: What Actually Works
Strength training roughly halves overuse injuries and makes you more economical. The evidence, the dose, and how to know which muscle to train first.
Ask a room of runners what they changed after their last injury and you will hear about shoes, insoles, foot strike, a stretching routine, maybe a cadence tweak. You will hear about strength training much less often, which is strange, because of everything on that list it is the only one with a large, consistent, randomised evidence base behind it. Strength training for runners is not a supplementary nice-to-have that serious people do in the off-season. On the current evidence it is the single most reliable thing you can add to your week if you want to stay healthy and run faster at the same effort.
This guide covers what the research actually shows, the dose that produced those results, why it beats the interventions runners usually reach for first, and how to work out which muscle group is your personal priority rather than doing a generic circuit and hoping.
Does strength training prevent running injuries?
Yes, and by a large margin. In the most-cited meta-analysis on the subject, strength training reduced sports injuries to less than a third of the control rate, and cut overuse injuries roughly in half. Overuse injury is the category almost every running injury falls into, so that second number is the one that matters to you.
That finding comes from Lauersen and colleagues, who pooled 25 randomised controlled trials covering 26,610 people and roughly 3,464 injuries. Their 2014 analysis in the British Journal of Sports Medicine found consistently favourable results for every prevention measure they tested, with one exception: stretching, which showed no protective effect at all.
They went back in 2018 with a tighter analysis of strength training specifically and found the same thing, with two useful additions. The effect was dose dependent, meaning more strength training produced more protection rather than plateauing early. And the interventions were safe and had high compliance, which is not something you can say about most injury-prevention advice.
Twenty-five randomised trials, 26,000 people. Strength training roughly halved overuse injuries. Stretching did nothing. That is the whole argument in two sentences.
Two honest caveats. Most of those trials were run in team sports and military populations rather than exclusively on distance runners, so the transfer is inferred rather than directly measured. And a meta-analysis tells you about averages across populations, not about you specifically. Neither caveat changes the direction of the finding, and nothing else on the runner’s intervention list has evidence anywhere near this strong.
Will lifting make me slower or heavier?
No. This is the objection that keeps runners out of the gym, and it does not survive contact with the literature.
Blagrove and colleagues reviewed the effect of strength training on the physiological determinants of middle and long distance running performance and found that heavy strength work, plyometrics, and combinations of the two improved running economy, which is the energy cost of holding a given pace. VO2 max and blood lactate were not improved, but critically they were not harmed either, and the strength-trained runners reached higher terminal velocities in maximal tests than endurance-only controls.
Running economy is worth being greedy about. It is the variable that separates two runners with identical engines, and unlike VO2 max it stays trainable deep into a running career. Getting a few percent more efficient means every pace you already run costs you less.
The bulk worry is mostly a misunderstanding of what heavy lifting does at the doses runners use. Two sessions a week of low-rep, heavy, mostly single-leg work, on top of your running volume, produces neural adaptation and tendon stiffness, not hypertrophy. You get stronger at close to the same bodyweight.
Why strength beats the things runners try first
Almost everything else on the standard list is either weak or a redirect.
Shoes. The idea that matching a shoe to your foot type prevents injury has been tested repeatedly and has not held up. Prescribing shoes by arch height or foot shape did not reduce injuries in military basic training trials, and a review of the whole practice concluded bluntly that prescribing distance running shoes on this basis is not evidence based. Shoes matter for comfort and for race-day performance, and comfort is a perfectly good reason to like a shoe. They are just not the injury lever they were sold as. We go through this in detail in do running shoes prevent injuries.
Stretching. No protective effect in the Lauersen data, and none in the dedicated reviews either. See stretching versus strength for what mobility work is and is not good for.
Form cues. Deliberately changing foot strike or “running tall” tends to revert within weeks and can shift load somewhere else rather than reducing it. The exception is cadence, which genuinely responds and genuinely changes joint loads.
Notice the pattern. The popular interventions all try to change the movement from the outside. Strength training changes the capacity underneath the movement, and the movement follows on its own.
What to actually train
Running is not a whole-body exercise in the way people assume. It is a very specific set of demands, concentrated in a small number of places, and that is where your gym time should go.
Work on muscle contributions during running showed that the ankle plantarflexors and hip muscles do most of the real work, with the emphasis shifting toward the hip as speed increases. That gives you a four-part priority list, and each item maps to a visible sign in your stride.
1. Glute medius and the lateral hip. Controls the pelvis and the knee in the frontal plane. Weakness here shows up as the hip dropping on one side or the knee diving inward. Both are the most reliable things a single camera can see. See contralateral pelvic drop and dynamic knee valgus. Hip-focused rehab has good evidence in patellofemoral pain, the most common running knee complaint.
Train it with: side plank with hip abduction, banded lateral walks, single-leg Romanian deadlifts, Copenhagen adductor holds for the other side of the same equation.
2. Glute max and hip extension. Your propulsion. Limited hip extension at toe-off is a common finding and usually means the glute is not contributing what it should. See reduced hip extension.
Train it with: hip thrusts, split squats, step-ups, hill sprints.
3. Soleus and the calf complex. The most under-trained tissue in distance running by a distance. The soleus absorbs and returns enormous force every single step, and it is heavily implicated in Achilles pain, calf strains and shin problems.
Train it with: heavy bent-knee calf raises for the soleus specifically, straight-leg calf raises for gastrocnemius, progressed to single leg and then to loaded. High reps to genuine fatigue, not fifteen easy ones.
4. Single-leg quad strength and trunk control. Running is a series of one-legged landings. Almost nothing in a bilateral squat rack tests that.
Train it with: rear-foot-elevated split squats, step-downs, single-leg squats to a box, and carries for the trunk.
A week that actually fits around running
Two sessions a week, 30 to 40 minutes each, is the dose most of the successful trials used, and it is enough. Three is better if you can protect the running.
- Session A (heavy, low rep): trap bar or conventional deadlift 3x5, rear-foot-elevated split squat 3x6 per leg, heavy bent-knee calf raise 3x8, side plank with abduction 3x30s per side.
- Session B (single leg and power): single-leg Romanian deadlift 3x8 per leg, step-up 3x8 per leg, pogo hops or skipping 3x20 contacts, Copenhagen plank 3x20s per side.
Rules that matter more than the exercise selection:
- Put it on a hard running day, after the run or several hours later, not the day before quality. Protect the running, not the lifting.
- Progress the load. The dose-response finding in the 2018 analysis is the practical takeaway: the same three sets of the same weight for six months is not strength training, it is a warm-up.
- Give it time. Strength adaptations run on a scale of eight to twelve weeks and up. Nothing in this guide works in a fortnight.
- Do not start a new strength block in race week. Everything else is negotiable.
Which one is your priority?
This is where most strength advice for runners falls down. It hands everyone the same circuit, when the entire point is that one of those four areas is your actual limiter and the other three are maintenance.
You can find your limiter two ways. The clinical way is a physio or S&C coach doing single-leg tests and watching you run. The at-home way is filming yourself and reading what your stride does under fatigue, which is what running gait analysis is for. The signal to trust is the left-right difference, not the absolute joint angle. A single phone camera carries five to fifteen degrees of error on absolute angles, but both of your legs are in the same clip at the same angle, so the comparison between them cancels most of that noise. Muscle imbalances in runners covers how to read that comparison.
Then train the weak side harder than the strong one, and retest in six to eight weeks. Retesting is the part everyone skips and the only part that tells you whether any of it worked.
What strength training cannot do
It cannot absorb bad load management. The most consistent single predictor of running injury is not weakness, it is a sudden jump in training. Novice runners who increased weekly distance by more than 30% in a week took on more distance-related injuries than those who stayed under 10%. No amount of squatting outruns a 40% mileage jump.
It also cannot treat pain you already have. If something hurts while you run, hurts the next morning, or is getting worse week on week, that is a physio or sports doctor conversation. Screening is not diagnosis, and a strength plan built on top of an undiagnosed problem is a slower route to the same place.
The short version
Shoes are comfort. Stretching is mobility. Strength is the one with the randomised trials behind it. Two sessions a week, heavy, mostly single-leg, biased toward the hips and calves, progressed over months, aimed at the side that is actually weaker.
That last bit is the hard part, and it is what RunGait is built for. It turns a short phone video of your run into a read on which of those four areas is limiting you, reports each signal with the confidence it genuinely deserves rather than inventing precision it cannot support, and hands you a four-week strength block aimed at the gap. Then you film it again and find out if you were right.