← All guidesMuscle Soreness: Does Being Sore Mean the Workout Worked?
It does not. Soreness measures how unfamiliar the work was, not how useful it was — it is not even an accurate read on how much damage you did. The one judgement worth learning here is the other one: sore, or injured.
Updated August 15, 2026
·
AI-researched
The answer, with evidence
Not lactate: A 1983 experiment pulled the two apart.
The same people ran for 45 minutes twice. Level running raised blood lactate significantly and produced no meaningful delayed soreness; running at a −10% grade never raised lactate at all and produced significant soreness. The dissociation holds in both directions. [1]
What does track it: Contracting while the muscle lengthens — eccentric work.
The DOMS review finds eccentric actions induce micro-injury at greater frequency and severity than other contraction types. The same review lists six competing mechanism theories and concludes that an integration of two or more is likely, so "your fibres tore" claims more than the evidence does. [2]
As a measuring stick: It does not even track how much damage you did.
110 men performed 12, 24, or 60 maximal eccentric actions. The 24 and 60 groups were significantly worse on every damage marker and recovered more slowly, yet their soreness on palpation and on flexion did not differ from the 12 group; soreness correlated with the other markers at r < 0.32. [3]
Still less results: Not being sore usually means the adaptation happened.
One bout of eccentric work protects against damage from the next — the repeated-bout effect, for which no unified mechanism has been established. [4] And damage is not the process that drives growth: protocols causing little of it produce comparable hypertrophy and strength gains. [5]
To feel less sore: What you can change is the feeling, not the repair.
Stretching — before, after, or both — reduced soreness by under 4 points on a 100-point scale, which the review calls not clinically important. [6] Massage, cold-water immersion, and compression do lower soreness scores, massage the most. [7] But those studies measured soreness ratings and blood markers, not strength or size weeks later.
Two treadmill runs separated lactate from soreness
"Lactic acid build-up" is the most durable explanation in the gym, and a clean 1983 design took it apart. The same subjects ran for 45 minutes twice, once on the level and once at a −10% grade. The level run pushed blood lactate up significantly and produced no meaningful delayed soreness. The downhill run never elevated lactate, and soreness clearly appeared over the following days. [1]
The force of that study is not its size but the fact that the dissociation runs both ways: lactate without soreness, and soreness without lactate. Any account that makes lactate the cause has to explain both halves, and it cannot.
What replaces it is mechanical: contracting a muscle while it lengthens produces microscopic damage to fibres and connective tissue, followed by inflammation and sensitised pain receptors. Worth keeping straight is that this is the leading account rather than a settled one — the review lists six hypotheses in all (lactic acid, muscle spasm, connective tissue damage, muscle damage, inflammation, and enzyme efflux) and concludes that two or more of them in combination is the likely explanation. [2] The part that is firm: eccentric actions induce micro-injury more often and more severely than other contraction types.
A prediction you can check
If eccentric loading is the key, the sorest movements should be the ones that load you while lowering or hold you at long muscle lengths: downhill running, slow squat and bench descents, Romanian deadlifts, Bulgarian split squats. The reverse should hold too — concentric-dominant work such as pushing a sled barely makes you sore. No study has ranked movements one by one, but the direction follows the review, and it will match your own record.
Soreness misses damage, and damage misses growth
Reading soreness as a score is a two-step chain: soreness means damage, damage means the session counted. Both links break.
The first link: 110 men performed 12, 24, or 60 maximal eccentric actions of the elbow flexors. The 24 and 60 groups were significantly worse on every damage marker and recovered more slowly — yet their soreness on palpation and on flexion was no different from the group that did 12. Correlating soreness against the markers put every coefficient below 0.32. The authors conclude that DOMS is a poor reflector of damage and inflammation. [3]
The second link: damage is not the mechanism of growth either. A review tracking the early weeks of resistance training reports that the first sessions’ increases in cross-sectional area are largely damage-induced swelling, with true hypertrophy appearing only after roughly 18 sessions — and, more to the point, that protocols causing little damage produce hypertrophy and strength gains comparable to protocols that cause it. Its conclusion is blunt: muscle damage is not the process that mediates or potentiates resistance-training hypertrophy. [5]
To be exact: nobody has run soreness against long-term gains head to head. The paragraph above composes two findings — soreness comes apart from damage, damage comes apart from growth — into an inference. Even granting only the first step, soreness has already stopped working as a metric.
It also explains why a few weeks on the same programme stop making you sore. A single bout of eccentric work is enough to protect you from damage in the next one — the repeated-bout effect. Its mechanism remains unsettled after decades of candidates, neural, mechanical, and cellular. [4] The phenomenon itself is not in doubt. Not being sore is what adaptation looks like, and your loads and reps are climbing across those same weeks.
The practical cost is concrete. To keep producing soreness you keep changing exercises, and changing exercises is what breaks progressive overload: you cannot add weight to a movement you replace every week. That is a practical argument rather than a research finding — and it is what the myth actually costs.
Is this soreness, or an injury?
This is the judgement to take away. One reason it deserves care is in the review: soreness reduces joint range of motion, shock attenuation, and peak torque, and it alters muscle sequencing and recruitment so that ligaments and tendons take unaccustomed stress — which is why returning to full intensity too early raises the risk of a further injury by itself. [2] What follows are observable differences, not a diagnosis. If anything in the right-hand column fits, stop and get it looked at.
What to look at
Ordinary soreness
Stop for this
Onset
Builds after training, worst at 24–48 h
Sudden, during the session, with a moment you can point to
Quality
Dull, diffuse, worse on pressure and movement
Sharp, stabbing, burning, or electric
Location
Across the muscle, usually both sides
One point, or inside a joint; one side only
Range of motion
Stiff, but loosens as you move
Genuinely restricted, weak, locking, or unstable
Trajectory
Steadily better over two or three days
No better after a week, or getting worse; swelling or bruising
Rare, but do not sit on it
If a session is followed by extreme soreness with visible swelling and your urine turns dark — cola or strong-tea coloured — get medical care that day rather than pushing through. That can indicate rhabdomyolysis. It is uncommon, and it is not the lifting-only problem it is described as: across a systematic review of 25 studies and 772 cases, 54.3% followed running and marathons against 14.8% from weightlifting, in mostly young men (mean age 28.7). The review’s own recommendation is to screen anyone presenting with muscle soreness, cramps, or dark urine after heavy endurance work. [8]
What you can change is the feeling, not the repair
Start with the one that does not work, because it is the one most often recommended. A Cochrane review pooled 12 randomised studies: stretching before exercise reduced next-day soreness by an average of 0.52 points on a 100-point scale (95% CI −11.30 to 10.26), and stretching afterwards by 1.04 points (95% CI −6.88 to 4.79). One large trial of 2,377 people stretching both before and after cut peak soreness across the week by 3.80 points (95% CI −5.17 to −2.43) — statistically significant, and far too small to matter. The conclusion: stretching produces no clinically important reduction in soreness. [6]
Some things genuinely do lower soreness ratings. A meta-analysis of 99 studies found massage, cold-water immersion, contrast water therapy, compression garments, cryotherapy, and active recovery all reduced DOMS to varying degrees (g from −0.40 to −2.26), massage most of all, with small drops in creatine kinase and inflammatory markers as well (SMD −0.37 for CK, −0.36 for IL-6, −0.38 for CRP). [7] Note what those studies measured: soreness ratings, perceived fatigue, and a few blood markers — not strength, size, or performance weeks later. The honest claim is "more comfortable", not "better trained".
The one lever that actually reduces how sore you get is in the review’s own practical advice: introduce novel and eccentric work progressively across a week or two rather than taking it to the limit the first time. [2] The rest is how to keep training while sore —
Lower the intensity and duration for the next day or two instead of cancelling. That is the review’s own advice for people who train daily. [2]
Train the less affected body parts. The sore muscles rest while the rest of you works — which is why "sore" does not mean "rest day". [2]
Keep moving lightly. Exercise is the most effective way to relieve pain during DOMS, but the analgesia is temporary and it does not speed repair. [2]
If soreness is degrading your movement quality, drop the load or the sets. Training through broken form is exactly the further-injury risk from the section above.
Sleep and protein are the parts that genuinely participate in repair — they just will not make you less sore today.
That last one is a whole guide of its own — see
the sleep guide. Whether this week needs a day off is a different question, answered in
the recovery guide — which deliberately leaves soreness off its under-recovery checklist, for the reasons above.
Keep the movement, move the load
Both practical conclusions on this page point at the same habit: the repeated-bout effect only pays off if the next session repeats the movement, and progressive overload only happens if what changes is the load or the reps rather than the exercise. The weekly plan in Omini is there to hold that side still — keep the movement, move the load.
Plan a training week
Frequently asked questions
Is muscle soreness caused by lactic acid?
No, and one 1983 experiment pulled the two apart in both directions at once. The same subjects ran for 45 minutes on the level, which raised blood lactate significantly and produced no meaningful delayed soreness, and again at a −10% downhill grade, which never raised lactate at all and produced clear soreness over the following days. The leading account of what does cause it is mechanical damage from contracting under lengthening load plus the inflammatory response to it — though the standard review lists six competing theories and concludes the real mechanism is probably a combination of two or more.
Does being sore mean the workout worked?
No, and soreness is a worse measure than most people expect: when 110 men performed 12, 24, or 60 maximal eccentric actions, the higher-volume groups were significantly worse on every damage marker yet reported no more soreness on palpation, with correlations below r = 0.32. Damage in turn is not what drives growth — protocols causing little of it produce comparable hypertrophy and strength. Soreness also fades on a programme you repeat, through the repeated-bout effect, while your numbers keep climbing. Judge by load, reps, and measurements over weeks.
How do I tell muscle soreness from an injury?
Ordinary soreness builds after training and peaks at 24 to 48 hours, feels dull and diffuse, sits across the muscle on both sides, loosens as you move, and improves steadily over two or three days. Stop and get it assessed if instead the pain appeared suddenly during the session, is sharp or stabbing, is localised to one point or inside a joint, affects one side only, comes with swelling, bruising, weakness, or genuine loss of range, or is not improving after a week.
Can I train while sore?
Usually yes, provided your movement quality holds — and you can train other muscles as normal regardless, which is why being sore does not by itself mean you need a rest day. If soreness is degrading your technique, reduce the load or the number of sets rather than cancelling the session. Low-intensity movement while sore generally feels better and does not interfere with repair.
How do I prevent DOMS? Does stretching or massage help?
Separate preventing it from relieving it. Stretching does neither: a Cochrane review of 12 trials found reductions of about half a point to four points on a 100-point soreness scale, which it calls not clinically important. Massage, cold-water immersion and compression genuinely do lower soreness ratings — massage most of all in a meta-analysis of 99 studies — but those trials measured soreness and blood markers, not strength or size weeks later, so the honest claim is comfort rather than faster repair. The one lever that reduces how sore you get is progression: introduce a new or heavily eccentric movement gradually over a week or two instead of taking it to your limit the first time.
When is soreness a medical emergency?
Rarely, but one combination should not wait: extreme soreness with visible swelling plus dark, cola-coloured urine after a session warrants same-day medical care rather than pushing through. That pattern can indicate rhabdomyolysis. It is uncommon, and it is not a lifting-only problem — across a systematic review of 772 cases, 54.3% followed running or marathons against 14.8% from weightlifting. This guide is general information and not medical advice.
References