Ice baths: what the evidence supports, at what dose, and when to skip them

The cold plunge has finished its journey from fringe to furniture. It is now standard equipment from London to Hong Kong — in purpose-built bathhouses, in gym recovery rooms, and in the corner of enough apartments that installing one has become a business of its own.

The interesting question is no longer whether cold water does anything. It clearly does. The question is what it does, at what dose, and what it costs you on the days you use it.

The short version: cold immersion is a good tool for feeling recovered and a poor tool for becoming stronger. Those two goals are in direct tension, and which one you are serving on a given evening should decide whether you get in.

What it reliably does

It reduces soreness and markers of muscle damage. This is the best-supported effect. A 2025 network meta-analysis in Frontiers in Physiology compared doses of cold water immersion for recovery from exercise-induced muscle damage and found clear winners. For delayed-onset muscle soreness, 10 to 15 minutes at 11 to 15°C ranked best (SUCRA 84.3 per cent). For creatine kinase — a blood marker of muscle damage — and for jump performance, 10 to 15 minutes at 5 to 10°C ranked best (SUCRA 75.7 and 70.4 per cent respectively).

Note what that table implies. The colder protocol is not better for everything. For soreness specifically, the merely cold water performed better than the very cold water, which is convenient, because 12°C is a great deal easier to sit in for twelve minutes than 6°C.

It lowers stress, but on a delay. A 2025 systematic review and meta-analysis in PLOS One pooled 11 randomised trials and 3,177 participants on health and wellbeing outcomes. Stress was significantly reduced 12 hours after cold immersion (SMD −1.00), with no significant effect immediately, at one hour, at 24 hours or at 48. Quality-of-life scores improved at 30 days and the effect had faded by 90. Sleep quality improved. Mood showed no significant difference against control. Cold showers were associated with a 29 per cent reduction in sickness absence.

That review also found something people rarely quote: inflammation increased immediately after immersion (SMD 1.03) and at one hour (SMD 1.26). Cold water is a stressor. The benefit, where it exists, is downstream of the stress, not instead of it.

What it appears to cost

If you lift to grow muscle, cold immersion straight afterwards is working against you.

A 2024 systematic review with meta-analysis in the European Journal of Sport Science — the one titled "Throwing cold water on muscle growth" — pooled 8 studies on post-exercise cold water immersion and resistance-training hypertrophy. The primary analysis favoured resistance training alone over resistance training plus immersion (cSMD −0.22, 95 per cent credible interval −0.47 to 0.04), with a 95.7 per cent probability that the true effect is negative.

Two things about that number, both of which matter.

It is a small effect, and its credible interval crosses zero. This is not evidence that an ice bath erases your training. It is evidence of a modest, consistent tax.

And the authors themselves flag that the underlying studies are of fair-to-poor quality. Anyone quoting this literature as settled — in either direction — is over-reading it.

The mechanism proposed is that cold blunts the acute inflammatory and anabolic signalling that follows resistance training, and that signalling is part of how the adaptation happens. That story is plausible and it is consistent with the strength findings, where routine post-lifting immersion has been associated with attenuated gains in maximal strength.

The decision, in one table

Situation	Plunge?	Dose
After a race or competition, season over	Yes	10–15 min, 5–10°C
Multi-day event, another race tomorrow	Yes	10–15 min, 5–10°C, as soon as practical
Congested fixture block, performance tomorrow matters more than adaptation	Yes	10–15 min, 11–15°C
Very sore, sleep being affected	Yes	10–15 min, 11–15°C
After a hypertrophy or max-strength session	No	Leave 4–6 hours, or skip
Base or build block, adaptation is the goal	Mostly no	Reserve it for the worst weeks
Before a strength session	No	Cold muscle produces less force
Because it is Tuesday and the plunge is there	Optional	Treat as wellbeing, not recovery

The organising principle: cold immersion buys tomorrow at the expense of next month. When tomorrow genuinely matters more — a second race day, a tournament, a week where you cannot afford to be wrecked — it is the right trade. When next month matters more, it is not.

If you want the benefit without the tax, the simplest lever is time. Separate the cold from the lifting by four to six hours, or take it on a different day entirely. Almost all of the reported interference comes from immersion immediately after the session.

Doing it properly

Temperature. Somewhere between 5 and 15°C covers everything the evidence supports. Colder is not more virtuous. Most commercial plunges run somewhere between 3 and 10°C, which is at or below the cold end of the studied range, so shorten the time accordingly.

Duration. Ten to fifteen minutes at 11 to 15°C. Two to five minutes is plenty at 3 to 6°C. Anyone doing fifteen minutes at 4°C is no longer running a recovery protocol.

Immersion depth. Studies overwhelmingly use whole-body or lower-body immersion to the waist or chest. Sitting with your feet in cold water is a different, much weaker intervention.

The first minute is the hard one. The cold shock response — the gasp, the involuntary hyperventilation, the spike in heart rate — peaks in the first 30 to 60 seconds and then settles. The technique is simply to control the exhale until it passes. If your breathing is not under control by 60 seconds, get out.

Never in open water alone, and never after drinking. The cold shock response is the mechanism behind most cold-water drownings. In a controlled plunge with staff nearby this is a low-risk activity; in a reservoir at dawn on your own it is not.

Talk to a doctor first if you have a cardiac condition, uncontrolled hypertension, Raynaud's, or are pregnant. Cold immersion sharply raises blood pressure and cardiac workload in the first minute. This is the one contraindication worth taking seriously rather than nodding at.

Rewarm passively. Dry off, put clothes on, move gently. Core temperature keeps falling for a while after you get out — the afterdrop — so the shivering ten minutes later is normal and is not a reason to get back in.

What about the sauna next to it

Most venues now sell the pair. Whether you should go heat first or cold first, and whether contrast cycling does anything the individual exposures do not, is a separate question with a clearer answer than you would expect: see sauna or ice first.

If your interest in the sauna is performance rather than pleasure, the heat literature is the stronger of the two — sauna protocols for athletes covers the dose and the evidence.

The honest summary

Cold immersion makes you feel better tomorrow, reduces soreness, appears to help sleep and stress, and probably costs you a small amount of muscular adaptation if you take it immediately after lifting.

None of that makes it a magic intervention or a mistake. It makes it a tool with a known direction of effect, which is more than can be said for most of what gets sold as recovery.

Use it when the next day matters. Skip it when the next month does.