CO2 tolerance test
The version of this test that spread through performance coaching does not time a breath hold at all. You take four normal breaths, one full inhale, and then time how long you can let the air back out in a slow, controlled stream.
The timer here runs that protocol and places the result in the bands Huberman Lab publishes. Sit down before you start.
Opens in exhale mode already set
- Times a slow controlled exhale, not a breath hold
- Four normal breaths, then one full inhale
- Stops when your lungs are empty
- Scored against the published Huberman Lab bands
This is a different measurement from the parent tool, so it keeps its own history. Your CO2 baseline is untouched.
Press start, take four normal breaths, then read the instruction
You are timing an exhale, not a hold
This is the part people get wrong before they get anything else wrong. The clock starts when the exhale starts and stops when the lungs are empty. Holding your breath at the top adds seconds and measures nothing.
The published instruction is blunt about it: stop the timer when your lungs are empty, because this is not about how long you can hold your breath. A held pause partway through the exhale ends the attempt too.
Who actually built this test
Huberman Lab credits the protocol to Brian Mackenzie, a human performance coach, and it reached most people through the Huberman Lab podcast and the guest series with Andy Galpin. Calling it the Galpin test, as a lot of the internet does, gets the attribution wrong.
That matters because two different band tables circulate under the same name. This page uses the three-band table published on hubermanlab.com, which pairs each band with a box-breathing duration. The five-band 20 / 40 / 60 / 80 table sits behind an email form and has no citable public source, so it is not used here.
Nose or mouth changes the number
The two source descriptions disagree. One specifies a nasal exhale; the other describes breathing out through the mouth as if through a narrow straw. A nose-only exhale is slower, so the same person scores differently depending on which they use.
Pick one and never change it. Write down which you used, along with the time of day and whether you had eaten. Comparing an unrecorded nasal test with an unrecorded mouth test tells you nothing.
What the score does not mean
- It is not validated against anything. No peer-reviewed study supports these bands. The one study the protocol cites is unpublished and about anxiety.
- It does not measure carbon dioxide. The name is a label for the exercise, not a measurement of blood gas.
- It is not a health check. No band here indicates a condition, and nothing on this page is medical advice.
- It is not a competition. Use it as your own baseline, retested every week or two under the same conditions.
The published bands, and how this timer maps to them
Compiled 2026-09-04 from the band table published on hubermanlab.com; the timer above starts on the exhale and stops when you press the button, so the seconds it reports are the ones these bands are written for.
| Band | Label as published | Prescribed practice | What the timer reports |
|---|---|---|---|
| 25s or less | Low CO2 tolerance | 3-second box breathing, about 2 min | Exhale duration in whole seconds |
| 25 to 45s | Moderate CO2 tolerance | 5 to 6 second box breathing, about 2 min | Exhale duration in whole seconds |
| 45 to 50s | Not labelled in the source | Not specified | Exhale duration in whole seconds |
| 50s and over | High CO2 tolerance | 8 to 10 second box breathing, about 2 min | Exhale duration in whole seconds |
Three published bands, not four. The gap between 45 and 50 seconds is real: the source table ends its middle band at 45 and starts its top band at 50, and we have left the gap open rather than quietly widening a band to cover it.
The overall dosing published alongside the table is two to three minutes of box breathing, once or twice a week, with a retest after one to two weeks. The timer records to its own history so you can compare a retest against the previous attempt rather than against a band.
Where the mapping breaks: the timer counts whole seconds from the moment you press start, so a slow first second and a hesitant stop both land in the number. Near a band boundary that is enough to move you. Treat a 24 and a 26 as the same result.
Source: Huberman AD, Breathwork Protocols for Health, Focus & Stress, Huberman Lab Neural Network newsletter, hubermanlab.com — the test credited there to Brian Mackenzie.
Questions
- What is the CO2 tolerance test?
- Four normal breaths, one full inhale, then a slow controlled exhale timed until your lungs are empty. It is an exhale-duration test, not a breath hold, and it is credited to Brian Mackenzie.
- What is a good CO2 tolerance test score?
- The published table calls 25 seconds or less low, 25 to 45 seconds moderate, and 50 seconds or more high. It does not label the gap between 45 and 50. None of the bands are validated against any performance or health outcome.
- Do I exhale through my nose or my mouth?
- The source descriptions disagree, so pick one and use it every time. A nasal exhale runs slower than a mouth exhale, which means the two are not comparable against the same bands.
- Is this the same as the BOLT score?
- No. BOLT times a breath hold after a normal exhale and stops at the first urge to breathe. This times a controlled exhale after a full inhale. Different protocols, different numbers, different published bands.
- How often should I retest?
- Every one to two weeks, under the same conditions: same time of day, same position, same route in and out, no coffee or hard exercise beforehand. Anything more frequent is measuring the day, not the training.
- Does a low score mean something is wrong with me?
- No. The test has no validated relationship with health or performance, and nothing on this page is a diagnosis. Use the number as your own starting point and watch whether it moves.
Where to go next
- CO2 baseline test — the parent tool, which times a hold to the first urge to breathe
- BOLT score test — the other named protocol on this timer, with McKeown's published bands
- Box breathing — the practice the bands prescribe, at 3, 5 or 8 seconds a side
- CO2 tolerance explained — what the term covers, and the limits of what a breath test can show