The short answer: 10 to 15°C. That is the range Royal Life Saving Society Australia describes as "generally considered safe for most individuals" in its cold water immersion therapy position statement, current as at February 2024, and AUSactive gives the same band. It is also where the best available dose-response evidence sits. Colder is available on almost every chiller sold here. Colder is not better.
That last sentence is the one worth sitting with, because the entire equipment market is built around advertising minimum temperatures, and we sell a chiller that reaches 3°C. This article is about what the guidance and the research actually say, which is not the same as what a spec sheet implies.
What the Australian guidance says
Royal Life Saving Society Australia's position statement on cold water immersion therapy states that a temperature range of 10 to 15°C is generally considered safe for most individuals, notes that open natural water can be as low as zero degrees, and advises keeping the head and neck out of the water. AUSactive, the fitness industry peak body, publishes the same 10 to 15°C figure.
Neither publishes a hard duration limit. Both advise starting with shorter exposures and building gradually. Both flag raised cardiovascular risk for people with heart or circulatory conditions and recommend medical clearance first.
The absence of a minute-based cap is worth noticing rather than filling in. If you see a confident "three minutes maximum" online, it is not coming from either of these bodies.
What the research protocols actually used
The literature converges on a similar band, arrived at independently.
The Cochrane review by Bleakley and colleagues (2012) pooled 17 randomised trials with 366 participants and defined cold water immersion as water below 15°C. It found reduced muscle soreness against passive rest, with a standardised mean difference of −0.55 at 24 hours and −0.93 at 72 hours, while stating plainly that the trials were small and of low quality and that there was no evidence cold water beat contrast or warm-water immersion.
More useful for choosing a temperature is the network meta-analysis by Pan and colleagues published in Frontiers in Physiology (2025), covering 55 randomised trials and 1,139 participants. It compared six dose combinations spanning 5 to 20°C and under 10 to over 15 minutes, and found 10 to 15 minutes at 11 to 15°C best for perceived soreness, with 10 to 15 minutes at 5 to 10°C best for creatine kinase and neuromuscular markers. Heterogeneity across the included trials was high.
So the answer depends slightly on what you are after, and both answers land inside or just below the guidance band. Neither points anywhere near 3°C.
Why colder is not better
The evidence does not reward it
No study in this literature found that going below 5°C produced better outcomes than the 5 to 15°C range. The dose-response work stops where it stops because that is where the trials were run, and extrapolating past the tested range is an assumption rather than a finding.
Colder water shortens the session
At 3°C most people cannot stay in for ten minutes, and the dose data associates 10 to 15 minutes with the better outcomes. A temperature you can tolerate for the researched duration is more defensible than one that drives you out at ninety seconds.
The cold shock response gets sharper
Sudden immersion in cold water triggers an involuntary gasp, hyperventilation and a jump in heart rate. Datta and Tipton described this response in the Journal of Applied Physiology (2006), and Shattock and Tipton set out the "autonomic conflict" hypothesis in The Journal of Physiology (2012): cold shock driving the heart faster while the diving reflex simultaneously drives it slower, which in susceptible people can trigger dangerous arrhythmias.
That literature comes largely from open-water and sudden-immersion settings rather than a controlled seated plunge at home, and the underlying mechanism is the same one. Colder water makes the response stronger. Entering slowly and breathing out as you go in is the standard mitigation, and it works.
Habituation is real but limited
Barwood and colleagues have shown the cold shock response habituates over roughly three to five exposures. Janský and colleagues (1996) found that repeated hour-long immersions at 14°C, three times a week for six weeks, produced no further sympathetic adaptation beyond that. Getting used to it is real; it is also not unlimited, and it is not a reason to keep dropping the thermostat.
What to actually set it to
Starting out: 15°C, which is the top of the guidance band, for two to three minutes. That is cold enough to be unmistakably cold and forgiving enough to let you learn your own breathing.
Settled in: 10 to 15°C for 10 to 15 minutes, which is where both the guidance and the dose-response data sit.
If you are chasing the biomarker end: 5 to 10°C for 10 to 15 minutes, per the Pan analysis. Below 10°C, staying in for the researched duration becomes genuinely hard for most people, which is a practical argument against it as much as anything.
Below 5°C: there is no evidence base recommending it and no guidance supporting it. If you go there, it is a personal preference, not a protocol.
Timing matters more than temperature for one group
If you lift for size or strength, when you plunge matters more than how cold it is.
Roberts and colleagues ran a 12-week randomised trial in 21 active men training twice weekly, comparing 10 minutes of cold water immersion at 10°C against active recovery after each session, and reported in The Journal of Physiology (2015) that strength and muscle mass increased more in the active-recovery group. Type II fibre cross-sectional area rose 17% and myonuclei per fibre 26% in that group only, and cold water blunted satellite-cell activation and p70S6K signalling.
One laboratory, 21 participants, not independently replicated by another group, and no meta-analysis on this question could be located. It is nonetheless the best evidence available, and it points one direction. Plunge on rest days, or hours away from a lifting session, rather than straight after it.
On our own product, plainly
The Ice Drop Chiller reaches 3°C. That is below the recommended band, and I want to be direct about what it means: it is a capability, not a recommendation.
A chiller with headroom holds 12°C effortlessly on a hot February afternoon, which is the practical reason a low minimum is worth having. It is not an instruction to run at the bottom of the dial, and a product that could only just reach 15°C would struggle to hold it when the room is warm.
Most of the time, on most days, you should be setting it somewhere between 10 and 15°C.
Temperature at a glance
- Royal Life Saving Society Australia and AUSactive: 10–15°C generally considered safe for most individuals
- Best-supported dose for soreness: 10–15 min at 11–15°C (Pan et al., 2025)
- Best-supported dose for biomarkers: 10–15 min at 5–10°C (same analysis)
- Cochrane definition of cold water immersion: below 15°C (Bleakley et al., 2012)
- Beginners: 15°C, two to three minutes, entering slowly
- No Australian body publishes a maximum duration. Build up gradually
- Medical clearance first if you have a heart or circulatory condition
Frequently asked questions
What temperature should an ice bath be?
Between 10 and 15°C, the range Royal Life Saving Society Australia describes as generally considered safe for most individuals. A 2025 network meta-analysis of 55 trials found 10 to 15 minutes at 11 to 15°C best for perceived soreness, and 5 to 10°C better for biomarkers such as creatine kinase.
Is colder better for an ice bath?
No evidence supports it. No study in this literature found outcomes improving below 5°C, and colder water shortens the session well under the 10 to 15 minutes the dose data associates with better results. It also sharpens the cold shock response on entry.
How long should you stay in an ice bath?
The dose-response data points to 10 to 15 minutes at 5 to 15°C. Neither Royal Life Saving Society Australia nor AUSactive publishes a maximum duration, advising instead to start with shorter exposures and build gradually. Beginners should start at two to three minutes.
Is 3°C too cold for an ice bath?
It sits below the 10 to 15°C band the Australian guidance describes and below any temperature the research has tested for benefit. Most chillers can reach it, which is useful as headroom for holding a set temperature on hot days, rather than as a target to sit in.
What temperature should a beginner start at?
15°C for two to three minutes, entering slowly and breathing out as you go in. That is the top of the recommended band, cold enough to produce the response and forgiving enough to learn your breathing before the cold shock reflex dictates it for you.
Should you ice bath after lifting weights?
Probably not immediately. A 12-week randomised trial in 21 men found strength and muscle gains were greater with active recovery than with 10 minutes at 10°C after each session (Roberts et al., The Journal of Physiology, 2015). One small study, but it points one way: plunge on rest days instead.
Setting a temperature you can hold
A chiller earns its place by holding a number rather than reaching a headline one. The Ice Drop Chiller, made by Enhanced Human, an Australian recovery brand based in Adelaide, drops a probe into the water and holds a set temperature without plumbing, in a tub you already own.
Set it in the band the guidance describes. The full specification is on the Ice Drop Chiller product page, and the studies referenced here are listed on our science page.
Related reading