Running has two distinct breathing problems, and most generic running-breath content addresses only the first one. The first is breathing during the run — pace, depth, nose vs mouth, when to switch. The second is breathing after — slow breathing sold as a recovery tool. The evidence for the first is modest. The best trial of the second found nothing, and this page used to say otherwise.
During the run: nasal breathing through zone 2
Below threshold — easy aerobic running, conversational pace, zone 2 in heart rate terms — nasal breathing is a reasonable choice, but not for the reason usually given. The pitch is that it works as a pace governor: easy runs drift too hard, and a nose-only rule stops the drift. Rappelt et al. tested exactly that in 19 fit adults doing an hour of self-paced easy cycling. Nose-only breathing lowered ventilation, breathing frequency, oxygen uptake and, late in the session, blood lactate. It did not change how much time people spent in each intensity zone.
“However, it did not prevent participants from performing low-intensity training at higher intensities than intended.”
That was cycling, not running, and one session, not a training block. But it is the direct test, and the governor failed it.
The long-term case rests on less. Dallam et al. (2018)tested ten recreational runners who had already trained nose-only for an extended period. Breathing nasally, they reached the same VO2max and time to exhaustion as breathing orally, with less ventilation for the same oxygen and better economy at a hard steady pace. That shows habituated runners give up nothing at the top end. It does not show that switching produces a gain: there was no comparison group that didn’t switch.
Above threshold — tempo runs, intervals, races — mouth breathe unless you have adapted to nasal breathing over an extended period, as Dallam’s runners had. For people who haven’t, nose-only breathing costs capacity at the top end: in Mapelli et al.’s (2025) twelve healthy adults, it lowered peak oxygen uptake, peak ventilation and peak workload in a maximal test, while changing little at rest or at the anaerobic threshold.
Breath-to-stride ratios
The traditional 2:2 ratio means two steps per inhale, two per exhale. At a cadence of 160 steps a minute, that is 40 breaths a minute. It’s fine. It’s not magic. It is also what most runners fall into: Bramble and Carrier (1983)recorded human runners using several breath-to-stride couplings (4:1, 3:1, 2:1, 1:1, 5:2 and 3:2) and found 2:1 the favored one. A stride there is a full gait cycle, two steps, so their 2:1 is the runner’s 2:2.
Asymmetric ratios — 3:2, 2:3 — get prescribed in running circles for injury prevention (alternating which foot strikes on the exhale). We have not found a controlled trial testing that claim. Use whatever ratio feels natural at your pace. If you’re bored on long runs, experimenting with breath patterns is fine; if you’re racing, the natural pattern is the right one.
After the run: slow breathing and recovery
This is where the marketing is furthest ahead of the evidence, and where this page used to be too. Raidl et al. ran the trial the claim needed: 34 people did four weeks of supervised sprint interval training, and those randomized to the breathing group also did slow-paced breathing after every session and every evening before bed. Performance improved in both groups, by amounts that did not differ. The authors found no physiological benefit from the breathing; the one between-group difference they report is that nocturnal heart rate rose slightly in the breathing group.
“These findings suggest that 10-minute sessions of slow-paced breathing following highly demanding sprint interval training offer limited value for enhancing performance or supporting recovery.”
Slow breathing does raise vagally mediated HRV during and right after a session — a 2022 meta-analysis of 223 studies, Laborde et al., found exactly that. The Raidl trial is the reminder that a change inside the session is not the same thing as faster recovery from training.
If you like five minutes of slow breathing while you stretch, do it for its own sake. Don’t count it as a recovery intervention. One trial of 34 people doing sprint intervals is enough to retire the confident claim. It is not enough to settle the question for distance runners.
HRV-guided training in practice
Read your HRV in the morning and compare it with your own recent baseline, not with anyone else’s. In Vesterinen et al.’s (2016)trial, 40 recreational runners either followed a fixed plan or did their moderate and hard sessions only on mornings when HRV sat inside an individually determined smallest-worthwhile-change band; on other mornings they ran easy. The HRV group did fewer moderate and hard sessions (13 on average, against 18), improved 3000 m performance by 2.1 percent where the fixed-plan group’s 1.1 percent was not significant, and the difference between the groups was small.
This is a brake, not an accelerator. The HRV signal doesn’t tell you to train harder when it’s elevated — it just gives you permission to. The veto on hard sessions is the part with evidence behind it: across eight studies, Düking et al. (2021) found that most HRV-guided programs prescribed fewer moderate or hard sessions and produced fewer non-responders, but their effect on performance and VO2peak was small and not statistically significant.
“However, if HRV-guided training is superior to predefined training for producing group-level improvements in fitness and performance, current data suggest it is only by a small margin.”
Training itself is the better-documented route to a higher resting HRV. Sandercock et al.’s (2005) meta-analysis found that exercise training increased high-frequency HRV power and RR interval, with smaller RR-interval responses in older people. The morning check is a way of reading that adaptation, not of producing it.
Wim Hof and other hyperventilation protocols
Wim Hof breathing — rounds of hyperventilation followed by a breath-hold at low lung volume — is popular in athletic circles. The acute response is real. In Kox et al.’s (2014) trial, volunteers trained for ten days in the method (the breathing, plus cold immersion and meditation) produced profoundly increased plasma adrenaline and a blunted inflammatory response to an injected bacterial toxin. The performance translation is a different matter.
“These findings indicate that despite large physiological effects, a single WHBM session does not improve anaerobic performance in repeated sprinting exercise.”
That was fifteen amateur runners, whose finger oxygen saturation fell to around 60 percent at the end of the breath-holds. A 2024 systematic review, Almahayni and Hammond, called the method’s effects on exercise performance mixed.
For an HRV-focused athletic application, Wim Hof and slow breathing are doing different things: one is built on hyperventilation and an adrenaline surge, the other is associated with a rise in vagally mediated HRV. Neither has shown a performance benefit for runners in the controlled trials above. Pick either for its own reasons, not for your race times.
For overtraining
If your seven-day HRV average has stayed low for a couple of weeks despite normal training, the fix is rest and a hard look at the training load. The ECSS/ACSM consensus statement on overtraining, Meeusen et al. (2013), frames it as excessive overload plus inadequate recovery, and found that none of the markers in use met the criteria for general acceptance. A low HRV trend is a reason to look, not a diagnosis.
No breathing protocol has been shown to shorten that recovery. What slow breathing can honestly offer in a rest week is ten quiet minutes a day. That is worth something. It is not a treatment, and it does not let you skip the rest.