Free UK & EU delivery on orders over £100
Ask five different cold plunge accounts what temperature to use and you'll get five different answers, usually delivered with total confidence and no citation in sight. The research is actually more specific — and more forgiving — than the "as cold as possible" advice you'll see repeated everywhere. Here's what the data supports.
When researchers set out to test cold water immersion for recovery, they don't use one standard temperature — they use a range, and that range is informative in itself. Across the trials pooled in Machado and colleagues' 2016 systematic review and meta-analysis in Sports Medicine, the water temperatures used across included studies clustered between roughly 11°C and 15°C (50–59°F), with immersion durations typically between 11 and 15 minutes.
This is the single most useful data point for anyone building a routine: the protocols that produced measurable reductions in muscle soreness were not using ice-water extremes of 3–5°C. They were using a firmly cold, but not brutal, range that most people can build a sustainable habit around.
This is where the meta-analysis gets genuinely useful, because Machado et al. specifically looked at whether varying the water temperature within the studied range changed the size of the soreness-reduction effect, and whether longer immersion times did either. Their analysis did not find a clear dose-response relationship where colder or longer reliably produced a bigger benefit within the ranges studied. In plain terms: dropping from 12°C to 8°C, or extending from 10 minutes to 20, is not obviously supported as "more effective" by the pooled evidence — it may just mean more discomfort for a similar result.
That's a meaningfully different message from most cold plunge marketing, which tends to imply that lower is always better. The honest reading of the evidence is: get into the range that's been studied, stay in it for a sensible window, and treat further intensity as a matter of personal preference and tolerance-building rather than a proven multiplier of the benefit.
Regardless of exactly where in the 10–15°C range you land, the physiology that matters most happens fast. Tipton and colleagues describe the "cold shock response" — the involuntary gasp reflex, rapid increase in breathing rate, and spike in heart rate and blood pressure that occurs in the first 30 to 90 seconds of cold immersion. This response is largely independent of fitness level and is the reason a controlled, gradual entry (and never plunging alone) matters more than shaving another two degrees off your set point.
The Šrámek immersion study is a useful companion data point here: it directly compared physiological responses across a range of water temperatures and found that colder water produced a substantially larger noradrenaline response than milder temperatures. That's part of why very cold water feels so bracing and alert-inducing — but it's the same mechanism underlying the cold shock response Tipton describes, which is precisely why easing in gradually (rather than jumping straight to the coldest possible setting) is the sensible way to build tolerance.
This article is for general information and does not constitute medical advice. Cold water immersion is not a medical treatment. Consult a professional if you have cardiovascular conditions.
A clear-eyed look at the evidence behind cold water immersion — muscle soreness, mood, sleep, and where the science is still thin.
The research clusters around a specific time window, not 'as long as you can stand it.' Here's what the meta-analysis found, and how to build up to it safely.
Cold showers and full immersion aren't the same intervention. Here's what the trial evidence says each one actually does, and which one fits which goal.
It's worth asking why the trials in the Machado meta-analysis converged on roughly 11–15°C in the first place, rather than something colder or milder. Part of the answer is practical: researchers need a protocol that participants can actually tolerate for the full immersion window without early withdrawal, since a trial where half the participants get out after 90 seconds doesn't produce usable soreness data. Part of it is physiological: this range is cold enough to produce meaningful vasoconstriction and a real drop in tissue temperature, without pushing straight into the more extreme end of the cold shock response that Tipton's review describes, where the risk of an involuntary gasp-and-panic response is highest.
That combination — cold enough to matter, tolerable enough to sustain — is a reasonable practical target for anyone building a home routine, independent of the research context. It's also worth noting what the meta-analysis's null finding on a dose-response relationship implies for equipment: if 8°C for 20 minutes isn't shown to outperform 12°C for 12 minutes, there's no evidence-based reason to chase colder specs on a chiller unit for their own sake. What matters more is being able to reliably land in and hold the range that's actually been tested.
Product marketing across the cold plunge category tends to lead with how cold a unit can go — some chillers advertise down to 3°C or lower. Against the evidence above, that headline number matters less than two other things: how consistently the unit holds a chosen setpoint within the 10–15°C range session to session, and how it performs as ambient temperature and body heat load the water during a plunge. A chiller that can technically reach 3°C but overshoots or drifts several degrees during a 12-minute session is harder to use for reproducing the studied protocol than one that holds 12°C tightly. If you're comparing tubs, ask about setpoint accuracy and recovery time between sessions, not just the minimum achievable temperature.
This is also where an external, calibrated thermometer earns its place in a routine even if your tub has a built-in display: built-in sensors can drift or read the water near the heating element rather than where you're actually sitting, and a two-or-three-degree discrepancy is enough to put you meaningfully outside the studied range without your realizing it.
A common failure mode for people starting out is treating the coldest tolerable temperature as the goal, then quitting a few sessions in because the experience is unpleasant enough to avoid. Given that the meta-analysis didn't find colder-is-better within the studied range, there's no evidence-based cost to starting at the milder end (15°C) and spending several weeks there before working down toward 10–12°C. What you gain from that gradual approach isn't a bigger physiological effect — it's a routine you'll actually sustain long enough to get the accumulated benefit of doing it consistently, which matters more than optimizing any single session.
The same logic applies to duration. Since the trials cluster around 10–15 minutes without showing that longer sessions do more, there's little reason to treat time in the water as a personal challenge to be extended indefinitely. A consistent 10–12 minutes at a temperature you can hold your composure in beats 20 minutes of clock-watching discomfort at a temperature you're gritting your teeth through.
Putting the research together into something you can actually use:
If you're deciding between a passive ice-bath tub and one with a chiller, the temperature research above is a good filter: you want a system that reliably holds somewhere in the 10–15°C range for the duration of a session, not one that starts there and drifts up by several degrees as the ice melts. A digital thermometer is the cheapest way to find out which situation you're actually in.
For the fuller picture of what the research says cold water immersion does and doesn't do — beyond just temperature — see Are ice baths good for you? What the research actually says.
None of the guidance above is useful if you don't actually know what temperature your water is at when you get in. Ambient room temperature, how long a tub has been sitting since its last chill cycle, and how much body heat previous sessions have added all shift the actual water temperature away from whatever the display last showed. The gap between "I set it to 12°C" and "the water was actually 12°C when I got in" is exactly the gap that determines whether you're replicating the studied protocol or quietly drifting outside it in either direction. A simple, calibrated digital thermometer closes that gap in seconds, and it's the cheapest way to know whether a discrepancy you're feeling is real or just perception after a long day.
This article is for general information and does not constitute medical advice. Cold water immersion carries cardiovascular risk, particularly for people with heart conditions, uncontrolled high blood pressure, or a history of arrhythmia. Consult a doctor before starting, never plunge alone, and stop immediately if you feel unwell.

Entry chiller for tubs up to 350L, 3–15°C range