Soreness is the most trusted feedback signal in training and one of the least reliable. What fifty years of research says it actually measures.
The session finishes. Two days later you can feel it walking down stairs, and something in you registers that as proof. The work landed. It counted.
Then a month goes by, you are training harder than you were, and the soreness has gone. The suspicion arrives quietly: has the training gone soft?
The research says the training is fine, and the reason has a name.
Soreness does not measure damage
The obvious assumption is that more soreness means more muscle damage, and more damage means more to rebuild. The first half of that fails immediately.
In a study of 110 men, participants performed 12, 24 or 60 maximal lengthening contractions of the elbow flexors. Strength loss, swelling, and creatine kinase in the blood all rose with the amount of work. Soreness did not. Twelve contractions and sixty contractions produced statistically indistinguishable soreness on palpation, and across every measure the correlation between soreness and actual damage was weak (Nosaka et al., Scand J Med Sci Sports, 2002).
Five times the work. The same ache.
The pattern is not confined to arms. When fifteen men trained nine different muscle groups with identical eccentric protocols in one session, soreness varied enormously by muscle. Elbow flexors, elbow extensors and spinal erectors reported the most. Legs, doing more total work, reported less (Chen et al., Scand J Med Sci Sports, 2019).
Think of soreness as a smoke alarm in one room of the house. It tells you something happened in that room and very little about the rest.
Damage is not what builds muscle
The second half of the assumption fails too.
Researchers tracked ten men through ten weeks of training, taking muscle biopsies at weeks one, three and ten. Muscle protein synthesis, the process that eventually builds tissue, was highest in the first week. So was muscle damage. And in that first week, the protein synthesis response showed no relationship to how much muscle anyone eventually gained.
Only later, once the damage had settled, did protein synthesis start correlating with actual growth (Damas et al., J Physiol, 2016).
The early response was repair: the body fixing what an unfamiliar stimulus had broken. Growth arrived once that repair work stopped being necessary. The review that followed put it directly. Muscle damage is not the process that mediates or potentiates training-induced hypertrophy (Damas et al., Eur J Appl Physiol, 2018).
What soreness tracks is novelty
Do something your body has not done recently and it will be sore. Do it again and it will not.
This is the repeated bout effect, one of the most reliably reproduced findings in the field. A meta-analysis of twenty studies using compound gym exercises found soreness and blood damage markers significantly lower after the second bout, at both 24 and 48 hours, with strength holding up better as well (Doma et al., JSCR, 2023).
The protection lasts. When the same eccentric session was repeated at six, nine and twelve month intervals, the six month group still showed reduced soreness and faster strength recovery. The effect faded somewhere between nine and twelve months (Nosaka et al., MSSE, 2001).
Read that the other way round. If soreness is your measure of a good session, you are measuring how unfamiliar the session was. Staying sore would mean constantly changing what you do, and constant change is a reliable way to go nowhere. Progress comes from repeating a movement long enough to load it heavier than last time. The repeating is exactly what removes the soreness.
How much to trust it
Most of this literature uses maximal lengthening contractions of the elbow flexors on an isokinetic dynamometer, performed by young men who have not trained before. It is a soreness generator, chosen because it produces a large, measurable response. Good for studying the mechanism. A poor description of your Tuesday.
Soreness itself is measured on a visual analogue scale, a person marking a line to describe a feeling. It is the best tool available and it is still subjective.
None of that rescues the original assumption. The finding that soreness fails to track damage has held since 2002. The finding that damage is not required for growth comes from biopsy data in people who trained for ten weeks.
What to watch instead
Soreness tells you something was new. On whether the session was productive, it says very little either way.
The measures that do tell you are unglamorous and countable. The weight on the bar. The reps at that weight. Whether you finished the session you were given. Whether you are still doing it in eight weeks.
A well-run program should make you less sore over time while the numbers keep moving. That is the adaptation doing its job.
References
Nosaka K, Newton M, Sacco P. Delayed-onset muscle soreness does not reflect the magnitude of eccentric exercise-induced muscle damage. Scandinavian Journal of Medicine & Science in Sports. 2002;12(6):337-346. DOI: 10.1034/j.1600-0838.2002.10178.x
Chen TC, Yang TJ, Huang MJ, et al. Damage and the repeated bout effect of arm, leg, and trunk muscles induced by eccentric resistance exercises. Scandinavian Journal of Medicine & Science in Sports. 2019;29(5):725-735. DOI: 10.1111/sms.13388
Damas F, Phillips SM, Libardi CA, et al. Resistance training-induced changes in integrated myofibrillar protein synthesis are related to hypertrophy only after attenuation of muscle damage. The Journal of Physiology. 2016;594(18):5209-5222. DOI: 10.1113/jp272472
Damas F, Libardi CA, Ugrinowitsch C. The development of skeletal muscle hypertrophy through resistance training: the role of muscle damage and muscle protein synthesis. European Journal of Applied Physiology. 2018;118(3):485-500. DOI: 10.1007/s00421-017-3792-9
Doma K, Matoso B, Protzen G, Singh U, Boullosa D. The Repeated Bout Effect of Multiarticular Exercises on Muscle Damage Markers and Physical Performances: A Systematic Review and Meta-Analyses. Journal of Strength and Conditioning Research. 2023;37(12):2504-2515. DOI: 10.1519/jsc.0000000000004628
Nosaka K, Sakamoto K, Newton M, Sacco P. How long does the protective effect on eccentric exercise-induced muscle damage last? Medicine and Science in Sports and Exercise. 2001;33(9):1490-1495. DOI: 10.1097/00005768-200109000-00011