Strength Training for Runners: What the Research Says and a Two-Day Program

Somewhere between one in five and four in five runners get hurt in a given year, depending on which study you read, and novice runners get hurt at more than twice the rate of experienced ones. Strength training is the one thing in a runner’s week with consistent research support for speed, and the best-supported candidate for durability, and it is also the thing most runners skip or do badly: a ten-minute bodyweight circuit after a run, light weights for high reps, planks. This guide covers what the trials actually show, how much and how heavy, what to skip, and a two-day program you can run in an apartment, a building gym or Central Park.

The short version

• Strength training improves running economy (how much oxygen you burn at a given pace) by roughly 2 to 8 percent in trained runners, and time-trial performance along with it, without hurting VO2max or adding body weight.
• Heavy loads and jumps work. Light weights for high reps have the weakest evidence. The best results come from heavy lifting combined with plyometrics.
• Across sports, strength programs cut injuries to less than a third in a meta-analysis of 25 trials. In runners specifically the evidence is thinner: one small foot-strengthening trial was positive; a ten-minute unsupervised routine tested on 720 NYC Marathon first-timers did nothing.
• Two to three sessions a week for 6 to 14 weeks or longer, progressing to heavy loads. That is the dose most positive studies used; putting them on easy run days is coaching practice.
• Stretching does not prevent running injuries. Core work alone has not been shown to make anyone faster.

What strength training does for running

Economy and race times

Running economy is the oxygen cost of holding a pace; improve it and the same effort buys a faster pace. A 2018 systematic review of 24 studies (469 middle and long distance runners, ages 17 to 45) found “good evidence” that strength training improves economy, time-trial performance and sprint speed, with economy improvements of roughly 2 to 8 percent versus control, and no effect on VO2max or body composition. Most programs ran two to three sessions a week for 6 to 14 weeks (Blagrove and colleagues, Sports Medicine, 2018). A meta-analysis of 16 studies put the pooled economy gain at about 3.9 percent, with explosive (jump) training at 4.8 percent and heavy weight training at 3.7 percent, both significant, and isometric work not (Denadai and colleagues, Sports Medicine, 2017).

The largest and most recent synthesis, 38 studies and 894 runners, looked at actual performance (time trials and time to exhaustion). High-load strength training produced a significant moderate effect; heavy lifting combined with plyometrics produced a large one; plyometrics alone did not reach significance. VO2max did not change. The authors rate the certainty of evidence very low to moderate (Llanos-Lagos and colleagues, Sports Medicine, 2024). A separate meta-analysis of 28 studies in endurance athletes (not only runners) reached the same ranking: maximal-force training beat other intensities, and performance improved with no change in VO2max (Berryman and colleagues, IJSPP, 2018).

The single cleanest trial: 17 well-trained runners, half of whom added half squats at 4 sets of 4 reps with a 4-rep max, three times a week for eight weeks. Running economy improved 5.0 percent, time to exhaustion at maximal aerobic speed improved 21.3 percent, 1-rep max rose 33.2 percent, and body weight did not change (Storen and colleagues, MSSE, 2008). And for honesty’s sake, the counterexample: in a trial of 19 well-trained female endurance athletes, the 11 who added heavy leg work twice a week for 11 weeks got 40 percent stronger and saw no change in economy or a 40-minute performance run (Vikmoen and colleagues, PLoS ONE, 2016). Strength work is a strong bet, not a guarantee.

Two caveats for anyone running a 4:30 marathon rather than a 2:30 one. Most of these studies used trained or highly trained runners, so the effect sizes will not transfer one to one. And gains take time: a 40-week study in competitive distance runners found economy significantly improved at both week 20 and week 40, again with no change in body mass, fat mass or lean mass (Beattie and colleagues, JSCR, 2017).

Injury

This is where the marketing usually outruns the evidence, so here it is in full. A meta-analysis of 25 randomized trials and 26,610 participants found strength training programs reduced sports injuries to less than a third (relative risk 0.315) and overuse injuries by about half (0.527). Stretching did nothing (0.963) (Lauersen and colleagues, BJSM, 2014). A follow-up of six trials found the effect was dose-dependent: every 10 percent increase in weekly strength volume was associated with a 4.3 percentage point lower injury risk (Lauersen and colleagues, BJSM, 2018). The catch: those participants were army recruits, adolescents and athletes across many sports, and only four trials fed the strength estimate. This is not a running study.

In runners, three trials matter. The most local one is a null result: 720 first-time NYC Marathon runners were randomized 12 weeks before the race to a self-directed 10-minute strength program three times a week (quads, hip abductors, core) or to nothing. Overuse injuries that stopped people from finishing: 7.1 percent with the program, 7.3 percent without (Toresdahl and colleagues, Sports Health, 2020). Nearly half the runners had some injury during training. A positive one: 118 recreational runners, 57 of whom did an 8-week foot and ankle strengthening course and then kept training remotely; over the next 12 months, the 61 controls were 2.42 times more likely to get a running injury (Taddei and colleagues, AJSM, 2020). And a trial of 433 recreational runners doing a twice-weekly strength and foam rolling routine for 18 weeks found no difference overall, but the minority who actually completed nearly every session did substantially better than controls (Desai and colleagues, Scandinavian Journal of Medicine and Science in Sports, 2023; figures reported in secondary coverage).

Read together, these say two things. A short, unsupervised bodyweight routine is not the dose that works. And a program only works if it gets done, which, for most people, means someone watching.

Tendons and bone

Running is high repetition and low load. Tendons stiffen in response to the opposite: a meta-analysis of 27 studies found loading above about 70 percent of maximum produced a large increase in Achilles and patellar tendon stiffness, while lower loads produced essentially none (Bohm and colleagues, Sports Medicine Open, 2015). Bone is similar: a 2021 comparison found male high-level runners had 9 percent lower lumbar spine bone density than non-athletes, a gap that was no longer statistically significant after adjusting for body mass, and cited an earlier finding of low lumbar density in 37 percent of 44 male runners (Herbert and colleagues, European Journal of Applied Physiology, 2021). Running loads the legs; squats and deadlifts load the spine.

For women runners, strength training is one part of bone health, not the main one. In a prospective study of 259 young female athletes, bone stress injury incidence was 10.8 percent overall and rose to 46.2 percent in those with three risk factors (12 or more hours of exercise a week, a leanness sport, and dietary restraint) (Barrack and colleagues, AJSM, 2014). Energy availability and menstrual function drive that risk; lifting does not fix under-fueling.

How much, how heavy, and when

• Frequency: two sessions a week. The reviews above found most positive programs used two to three. Three in an early base phase, two as mileage climbs, is a common coaching pattern and the one the Hospital for Special Surgery recommends for marathon builds (HSS, 2024).
• Duration: 6 to 14 weeks minimum, longer is better. Economy gains grew with program length in the 2017 meta-analysis, and the 40-week study still showed gains at week 40.
• Load: heavy, low reps. The heavy prescriptions in the 2018 review were 2 to 6 sets of 3 to 10 reps above 70 percent of 1-rep max. In the one trial of masters marathoners (16 runners in their mid 40s), a group lifting 3 to 4 reps at 85 to 90 percent improved economy at marathon pace by 6.17 percent; a group doing 3 sets of 10 at about 70 percent did not improve (Piacentini and colleagues, JSCR, 2013; loads from secondary coverage). Tiny study, but it points the same way as the meta-analyses.
• Add jumps. Six weeks of plyometrics improved 3 km time 2.7 percent in one trial of 17 male runners (Spurrs and colleagues, European Journal of Applied Physiology, 2003) and improved economy in a six-week trial of 18 recreational runners (Turner and colleagues, JSCR, 2003). Combined with heavy lifting, plyometrics gave the largest performance effect in the 2024 meta-analysis.
• Will it hurt the running? The 2022 meta-analysis of 43 concurrent-training studies found lifting plus endurance work did not compromise maximal strength or muscle size; only explosive strength was slightly blunted, and less so when sessions were separated by at least three hours (Schumann and colleagues, Sports Medicine, 2022). In the other direction, none of the reviews above found strength work reduced VO2max or endurance.
• Timing. Heavy lifting impaired running economy six hours later and again the next day in one small study (Doma and Deakin, 2013). The reviews disagree on whether to lift or run first on a double day; spacing matters more than order. Our practice, which is coaching convention rather than a tested threshold: lift on easy run days, after the run or at least a few hours apart, and never the day before a long run or a key workout.

Five things runners believe that the evidence does not support

• “Lifting will make me heavy.” Body mass did not change in the 8-week squat trial, the 40-week study or the masters trial above, and the 24-study review found body composition typically unaffected. Muscle fibers can grow without the scale moving (the Vikmoen trial found fiber hypertrophy with no performance change), so this is “the scale does not move,” not “you cannot gain muscle.”
• “Runners should lift light and high rep.” Maximal-force training beat other intensities in the 2018 meta-analysis; high load was the method with a significant performance effect in 2024; the light-load group in the masters trial did not improve. Light circuits are not useless, they are just not what moved race times.
• “Core work makes you faster.” No good trial shows it. The one often cited, 8 weeks of extra core sessions in 21 college athletes, improved plank endurance but its between-group effects were not significant and it did not measure race performance (Hung and colleagues, PLoS ONE, 2019). The NYC Marathon routine that did nothing was built around quads, hip abductors and core. Keep some trunk work; load the legs.
• “Stretching prevents injuries.” A Cochrane review found no evidence that stretching prevents lower-limb running injuries (6 trials, 5,130 participants, relative risk 0.85, confidence interval crossing 1) (Yeung and colleagues, Cochrane, 2011), and the 2014 sports meta-analysis agreed. Stretch if you like how it feels.
• “Weak hips cause runner’s knee.” A systematic review found hip abductor weakness was associated with iliotibial band syndrome in distance runners, but the link to patellofemoral pain was unclear (Mucha and colleagues, Journal of Science and Medicine in Sport, 2017). Hip work does help treat existing knee pain: adding hip strengthening to knee strengthening beat knee work alone across 14 trials (Nascimento and colleagues, JOSPT, 2018). Treatment, not cause.

The two-day program

This is the shape of what we build for runners, with the loads and reps as a common starting point rather than a prescription. It needs a pair of adjustable dumbbells or a kettlebell at home, or a rack and barbell in a building gym; the jumps go in the park. Both days take about 45 minutes. Put them on easy run days, after the run, and keep the day before your long run clear.

Day A: heavy legs

• Warm-up: 5 minutes easy, then leg swings, walking lunges, 10 pogo hops.
• Back squat or goblet squat: 4 sets of 4 to 6. Start at a weight you could do 10 of and add load weekly. By week 6 these should be heavy.
• Romanian deadlift (dumbbell or barbell): 3 sets of 6 to 8. Hinge at the hips, flat back, feel the hamstrings. See our deadlift vs squat guide for which to lead with.
• Standing calf raise, straight knee, slow lowering: 3 sets of 10 to 12, loaded (a dumbbell in one hand, the other on a wall). Progress to single leg. This is the one exercise most runners never load and the one the tendon research points at.
• Side plank: 3 holds of 20 to 40 seconds each side.

Day B: single leg and jumps

• Warm-up: as Day A.
• Bulgarian split squat or reverse lunge: 3 sets of 6 to 8 per leg, dumbbells in hand. A split squat loads one leg about as hard as a squat loads two, which matters when the heaviest thing in your apartment is a 40-pound dumbbell.
• Single-leg Romanian deadlift: 3 sets of 8 per leg, light, focus on balance and the hinge.
• Hip abduction (banded side steps or side-lying leg raise): 2 sets of 15 per side. For the ITB-prone.
• Seated or bent-knee calf raise: 3 sets of 12.
• Plyometrics, outdoors or on a tolerant floor: after 4 weeks of the strength work, add 40 to 60 foot contacts per session: pogo hops, skips for height, broad jumps, single-leg hops. Build to 100 or so over six weeks. See plyo box exercises for progressions.

Progression is the whole point. If the weight on Day A is the same in week 8 as in week 1, the research above does not apply to you. Feet and ankles get their own attention in our ankle-strengthening guide; hills do some of this work for free, see hill workouts; and if a heavy session leaves you flat, the next day is a recovery run, not a tempo.

If you run in Central Park or train for New York

Harlem Hill climbs about 84 vertical feet at a 4.4 percent grade over a third of a mile when you run the loop counterclockwise from 110th Street (Central Park Track Club). The NYC Marathon’s hardest climbing comes at miles 23 and 24 on Fifth Avenue, after 22 miles of running, according to a third-party course analysis (RunnersConnect). Both are arguments for leg strength that survives fatigue, which heavy lifting builds and which a 10-minute bodyweight circuit was never shown to. 59,226 people finished the 2025 race (NYRR, via Running USA); if you are aiming for 2027, our guide to getting into the NYC Marathon covers the routes in, and the strength work above should start now, not in July.

Masters runners

Distance running does not protect against age-related loss of power. In a study that followed 89 masters athletes aged 35 to 90 for about four and a half years, jump power and force declined at the same rate in sprinters and distance runners, but the power-trained athletes stayed ahead throughout (Ireland and colleagues, Calcified Tissue International, 2022). The masters marathoner trial above is small but points the same way: heavy loads improved economy at marathon pace in runners in their mid 40s, moderate loads did not. If you are over 45, the heavy day is the one not to skip.

What happened to the old post

This guide replaces “Strength Training for Runners: The Definitive Guide,” an 18-section article that now redirects here. The old version recommended 10 to 15 rep ranges and did not distinguish between what the trials found in trained runners and what applies to recreational ones. The advice has changed, not just the length.

Working with a trainer on this

Type A Training’s trainers run strength sessions for runners in apartments, building gyms and parks across Manhattan and parts of Brooklyn and Queens, built around your race calendar: heavier and three days a week in the base phase, two days and lighter as mileage peaks, and the plyometrics in the park. The research says the program works when it is heavy, progressive, and done for months, and the 433-runner trial says the people who got the benefit were the ones who showed up. That second part is what a standing appointment is for. See in-home personal training, or text (917) 951-0111.

Frequently asked questions

How often should runners strength train?

Most positive studies used two to three sessions a week; two is the practical minimum, three in a base phase if you can recover from it. Put them on easy run days and keep the day before a long run or hard workout clear.

Should runners lift heavy or light?

Heavy. One meta-analysis found maximal-force training beat other intensities, the largest found high load was the method with a significant performance effect, and the one masters trial found 85 to 90 percent loads improved economy while 3 sets of 10 did not. Build up to heavy over several weeks; do not start there.

Will strength training make me slower or heavier?

None of the reviews above found strength work reduced VO2max or endurance, and body mass did not change in the 8-week, 40-week or masters trials. Explosive strength is slightly blunted when lifting and running happen in the same session, so separate them by a few hours.

Does strength training prevent running injuries?

Across sports, strongly yes (injuries cut to less than a third in a 25-trial meta-analysis). In runners specifically the evidence is mixed: a foot-strengthening trial was positive, a 10-minute unsupervised routine tested on NYC Marathon first-timers had no effect. Progressive, loaded, consistent work is the version with the best odds.

Do I need a gym?

No. A pair of adjustable dumbbells covers split squats, Romanian deadlifts and calf raises at useful loads; a building gym with a rack lets the squat get properly heavy; the park is for the jumps.