How many times a week should you lift as a runner?

Two sessions a week can keep you getting stronger, as long as those two sessions carry enough hard sets. When studies matched weekly volume, how many days it was spread over stopped mattering for strength ¹. A newer analysis does find more days helps strength a bit on its own, with diminishing returns ². So think in sets per week first and days second, and keep your heavy leg work out of the same session as a run.

Frequency is mostly how you fit the sets in#

The paper most often quoted here is Grgic 2018, which pooled 22 studies. It's worth reading both halves, because only the first half usually gets quoted.

Strength gain by lifting frequency, volume NOT matched

	Value (effect size)
1x per week	0.74
2x per week	0.82
3x per week	0.93
4+ per week	1.08

Pooled effect sizes from Grgic et al. 2018 (22 studies). When weekly volume was matched, frequency had no significant effect (p = 0.421).

That rising pattern is real, and frequency had a significant effect overall (p = 0.003). Then the authors looked only at studies where weekly volume was equal across groups, the same number of sets spread over more or fewer days. The effect went away, p = 0.421. Their wording is that the frequency effects "seem to be primarily driven by training volume" ¹.

In plain terms, people training four days a week got stronger mostly because four days made it easy to do more sets.

Two caveats from the same authors. They say it "remains unclear" whether frequency matters on its own, and most participants were untrained, so they ask for more evidence in trained lifters. Only four of the 22 studies were rated good quality, the rest moderate ¹.

One subgroup result matters for runners. Frequency had a significant effect for multi-joint lifts (p < 0.001) and for upper-body strength (p = 0.004), but not for lower-body strength (p = 0.070) ¹. A non-significant result isn't proof that legs don't care how often you train them. It does make the case for a third leg day weaker than the case for a third upper-body day.

The newer evidence partly disagrees. Pelland and colleagues ran meta-regressions across 67 studies and 2,058 participants. Volume drove both size and strength, with diminishing returns that were "considerably more pronounced" for strength. Frequency did little for muscle size but did increase strength gains, again with diminishing returns ². It was first posted as a preprint in 2024 and is now published in Sports Medicine.

So frequency isn't irrelevant. It just isn't the main lever. If two days is what you've got, the question isn't how to find a third. It's whether the two you have do enough work.

How many sets#

Ralston 2017 grouped studies by weekly sets per exercise: 5 or fewer, 5 to 9, and 10 or more. Ten-plus beat five-or-fewer for strength, but by a modest margin. The difference was ES 0.18 (95% CI 0.06 to 0.30), with the low group at 0.82 and the high group at 1.01 ³.

Weekly sets per exercise and strength gain

	Low (effect size, pre to post)	High (effect size, pre to post)	Mark (effect size, pre to post)
Low weekly sets	0.47	1.17	0.82
High weekly sets	0.7	1.32	1.01

Mark = pooled effect, bar = 95% CI. Low = 5 or fewer sets per exercise per week, high = 10 or more. Source: Ralston et al. 2017 (9 studies, 223 men).

Both groups got clearly stronger. The extra sets bought something, not everything. That fits Pelland's diminishing returns for strength ². Ralston's evidence is thin, though: nine studies, 223 men, and only one of the nine was a randomised controlled trial. The authors also flag that repeated 1RM testing can inflate measured gains through practice alone ³.

Schoenfeld's grouping of under 5, 5 to 9 and 10 or more sets per muscle only showed a trend for size (p = 0.074), even though each extra set helped on the continuous measure ⁴. More is better on average, without a clean threshold where it starts to count.

Count carefully, because the papers don't count the same way. Ralston counts sets per exercise. Schoenfeld's 2017 hypertrophy analysis counts sets per muscle, and found each extra weekly set added about 0.37% to the muscle-size gain ⁴. Pelland goes further and counts indirect sets as half a set, which fit the data best ². A bench set is a full chest set and arguably half a triceps set. That's the kind of bookkeeping that decides whether your "12 sets" is 12.

Strength or size: why runners can lift less than bodybuilders#

Pelland's results split in a way that suits a runner. For muscle size, volume kept paying and frequency barely mattered. For strength, the returns on extra volume flattened much faster, and frequency helped ².

Our reading of that, and it is a reading rather than a tested plan: if you mostly want strength and don't need extra muscle to carry on your runs, you can get a lot of the benefit from moderate weekly sets spread over two or three exposures. Chasing high set counts buys mostly size, and size is extra mass to carry on every run, which running for lifters puts a number on.

The floor#

If you're short on time, this is the useful number. Androulakis-Korakakis 2020 asked what the minimum is for trained men to keep getting stronger. A single set of 6 to 12 reps at about 70 to 85% of 1RM, taken to failure, two or three times a week for 8 to 12 weeks, produced gains the authors call "suboptimal, yet significant". Pooled across five studies, 1RM went up 12.09 kg overall, 17.48 kg on the squat and 8.25 kg on the bench ⁵.

Read the conditions. Those sets were to failure. It was six studies, all in men, and the authors say it's unclear whether the same works for the deadlift, for women or for highly trained strength athletes ⁵.

So the floor is low. It isn't zero, and it isn't the same as optimal.

What counts as a set#

Two things trip people up here. The first is effort. The minimum-dose studies used sets taken to failure ⁵, and a set of eight with five reps left in the tank isn't what they tested. How close to failure is worth going when you also run is its own trade-off, covered in RIR, RPE and failure.

The second is running. None of the volume analyses above count a run as leg volume. Whether your weekly kilometres do anything for squat strength isn't something they answer. The concurrent-training data suggests running neither adds much nor takes much from maximal strength on average ⁶, so don't count it as a leg session and don't treat it as a reason to skip one.

What changes when you also run#

Running and lifting share legs, so the question is whether the running eats into lifting progress.

Overall, not much. Schumann 2022 pooled 43 studies and found adding aerobic training didn't compromise maximal strength or muscle size. How often people did concurrent training, more or fewer than five sessions a week, didn't change that ⁶.

The exception is experienced lifters and same-session training. Petré 2021 found concurrent training reduced lower-body 1RM gains in trained people (ES -0.35), not in untrained ones (ES 0.03). And in trained people the cost came from doing endurance and strength in the same session (ES -0.66). In separate sessions it was small and not significant (ES -0.10) ⁷. So the reassuring results from beginners don't carry straight over to someone who has lifted for years.

So the practical version for a runner: put the heavy leg session on its own, or at least a few hours away from a run. Of the things you control here, that's the one with the clearest evidence behind it.

A real check: my last six weeks#

Here's my own log since I got to campus in September, as a worked example of counting it properly.

Ken's working sets per week, autumn 2026

	Value (sets)
7 Sep	18
14 Sep	32
21 Sep	30
28 Sep	22
5 Oct	25
12 Oct	23

Working sets logged per week (Monday to Sunday), warm-ups excluded. Every set was upper body or core. Source: Ken's hybridAF log, read 2026-10-22.

That's 150 working sets across six weeks, on two to five lifting days a week, alongside 7 to 31 km of running a week. By Ralston's per-exercise count, pull-ups land in the 5 to 9 band in four of the six weeks, and barbell bench in two.

And there's the gap. Not one of those 150 sets is a squat, a deadlift or anything else for the legs. Plenty of lifting, plenty of days, and the half of the body that runs got none of it. Counting days would have hidden that. Counting sets per exercise shows it in one line.

My Strength Index has read 74 the whole time. That isn't evidence nothing faded, because the score grades my best logged barbell lifts rather than a recent window, so it wouldn't show a slow slide.

I said in a reel about the Grgic paper that I'd stopped counting sessions and started watching whether the number moved. This log is the footnote: watch what's in the sets too, because a number built on your best lifts won't tell you the legs have gone quiet.

Putting numbers on a runner's week#

None of the studies above prescribe a hybrid week. This is our reading of them, not a tested protocol.

Keep at least the lower-body sessions away from your hard runs, per Petré. The question of where those sessions sit in the week is covered in how to split a hybrid training week, and the very-low-time version in the minimum effective dose. If you're a runner deciding what kind of lifting to do rather than how much, start with does lifting make you a faster runner. The whole cluster is on the hybrid training hub.