Running Economy: Why Strength Work Makes You Faster Without More Miles

By Thomas Corvazier

September 7, 2026

Two runners with the same VO₂max rarely race the same time. A large part of the difference is running economy — how much energy you burn to hold a given pace. Better economy means the same speed costs you less.

Most runners try to improve it by running more. There is a cheaper lever, and the research on it is unusually specific about which kind of strength work pays.

Your pace changes the mix, not the answer

The 2024 review that anchors this article did something unusual: it reported results by the speed the runner was tested at.¹ That turns out to matter — but not in the way it is usually summarised.

Lifting heavy worked across 8.64–17.85 km/h. That span covers almost every runner, recreational and fast alike. So load is not the elite’s tool; it is everyone’s. What your pace changes is how much choice you have alongside it:

• Faster than about 12 km/h (quicker than 5:00/km, or 8:03/mile) — heavy lifting is the only method that kept working. Plyometric training stopped helping at those speeds, and trended slightly the wrong way. If you race here, load is not optional.
• At or below about 12 km/h — heavy lifting still works, and jumping works too. You have two levers instead of one, which also means you can get a real return without a gym. Load is still the stronger bet; it just is not the only one.
• At any pace in the studied range — doing both beat doing either. Combined training produced the largest improvement of any single method.

So this is not a beginner’s article or an elite’s article. It is the same evidence for both. The practical difference is that the faster you run, the more of the benefit has to come from the bar.

What the meta-analyses actually found

A 2024 systematic review in Sports Medicine pooled strength-training trials in middle- and long-distance runners and separated them by method. High-load and combined training produced small-to-moderate improvements in running economy. Plyometric training improved it at speeds up to about 12 km/h. And two methods produced no improvement at all against controls: submaximal-load training and isometric training.¹

Worth being precise about that last sentence, because “submaximal” is easy to misread. The categories are defined by load:

So the dead zone is moderate-weight lifting — the comfortable three-sets-of-ten middle ground. Bodyweight jumping is not in that dead zone: it is the plyometric row, which is why it earns a place below 12 km/h alongside the heavy work rather than instead of it.

“Do some strength work” is not the finding. Lift heavy, jump, or do both is the finding.

The part that surprises people: it does not work by making your legs bigger. The reviews attribute the gain to neuromuscular and musculotendinous changes rather than muscle size — a stiffer, springier leg returns more elastic energy per step and spends less time on the ground.² Programmed this way, body mass generally does not move.

Which muscles do which job

A stride is not one movement, and the muscles are not interchangeable.

Each section below gives two lists. Bodyweight is where you learn the movement and where to start if you have never lifted. Loaded is where the economy evidence actually sits — high-load work was the method that improved economy across the widest range of speeds, and “high-load” means above 80% of your one-rep max.¹ As a practical rule of thumb that is roughly the most you could lift about eight times before form fails, so it means few reps and real weight, not a comfortable set of fifteen. Bodyweight first, then load it.

Calves — the support, not the push

Up to around 7 m/s, the ankle plantarflexors — soleus and gastrocnemius — are the largest contributors to vertical support. They hold you up more than they push you forward, which is the opposite of how most people picture them.³

Bodyweight

Loaded

Glutes and hamstrings — the drive

Hip extension is what puts force behind you. Above ~7 m/s the hip muscles are also what raise stride frequency.³

Bodyweight

Loaded — the split squat is the one to prioritise: elevating the rear foot puts almost all of the load on one leg, which is how running works.

Quads — accepting the load

They absorb the landing rather than producing the push, which is why single-leg work matters more here than volume.

Bodyweight

Loaded — the guided versions are the easier place to go heavy, because balance fails before the legs do when you are holding dumbbells.

Trunk and obliques — stopping the leak

They keep the pelvis from dropping, so the force you produce goes into the ground instead of sideways.

Two or three short sessions a week is enough to matter. None of this replaces running; it makes the running you already do cost less.

One honest note on the trunk work above: isometric training was one of the two methods that did not improve economy in the review.¹ Those moves are here because holding the pelvis level is a real job during a stride — not because a plank buys you economy. Keep them; do not expect them to do the work of the jumps below.

The jumping half — plyometrics

This is the half most runners skip, and below 12 km/h it has real evidence behind it — as an addition to the heavy work, not a replacement for it.

What it actually means. In the review, plyometric training is defined as light-load work — under 40% of your one-rep max — done explosively, with the aim of training the stretch-shortening cycle and muscle–tendon stiffness.¹ In plain terms: your leg is a spring, and this trains the spring rather than the motor. Most of it is bodyweight.

What the trials actually did. The eleven plyometric studies used, in various combinations:

• Hops — single-leg hops, calf/ankle jumps
• Jumps — squat jumps, countermovement jumps, box jumps, hurdle jumps, depth (drop) jumps
• Bounds — single- and double-leg bounds, alternate-leg bounds, split scissor jumps

Typical dose: 6–12 weeks, 1–2.5 sessions per week, 2–5 sets of 4–20 repetitions, with 1–3 minutes rest between sets.¹ That is a small amount of work — it is closer to ten focused minutes than to a gym session.

Why the 12 km/h ceiling exists. The effect is not a quirk of the statistics; it has a mechanical reading. Plyometric work changes tendon stiffness and how much elastic energy the leg stores and returns on each contact. At slower speeds there is a lot of pre-stretch to exploit, so a springier leg returns more free energy per step. As speed rises, the tendon is already doing most of that job, so adding compliance stops helping — and in the trials the effect at speeds above 12 km/h was not merely absent but slightly unfavourable.¹

Two caveats worth stating plainly. Pooled across all speeds, plyometric training showed no significant effect — the benefit only appears once you separate the slow speeds out, and the review rates the certainty as moderate.¹ And jumping is impact work: build the volume up over weeks rather than starting at the top of those ranges.

References