The session
- Compromiser
- One station, the same one every session through a training block. Burpee broad jumps by default.
- Work
- 90 seconds on the compromiser, continuous.
- Transition
- None. You start running the moment the station ends.
- Run
- 5 minutes, at an effort you can repeat on every round.
- Rest
- 90 seconds after each run.
- Rounds
- Four to six. Start at four.
- Target zone or pace
- Neither. The run is governed by repeatability: the first rep sets the effort and every rep after it matches.
- Total time
- About 50 minutes at four rounds and about 70 at six, including a 10-minute warm-up and a 10-minute cool-down.
The 4-to-6 round range, the absent transition, and the swap rule for the compromiser are coaching consensus. The 90 seconds, the 5 minutes, the rest, the starting round count, and the default station are RunBeat's picks inside what those sources document. Each section below gives the reason for the pick.
What does a compromised running workout look like?
90 seconds on one station, straight into a 5-minute run, then 90 seconds of rest, four to six times through. The handover from station to run is the one place in the session where rest isn't allowed, and it's the reason the session exists. The compromise is produced by the absence of a transition.
Pick one station and keep it for the whole block. That constancy is what makes the session readable: with the compromiser fixed, a change in the running is a result rather than noise. What to track from session to session is covered in how should you use heart rate in compromised-run training?
The rest goes after the run rather than before it. 90 seconds is long enough to do the next round of station work properly and short enough that heart rate hasn't come back down. Every round after the first starts already compromised, which is the race condition reproduced on purpose.
At four rounds the whole thing takes about 50 minutes with a 10-minute warm-up and a 10-minute cool-down. At six it's about 70.
Why 90 seconds of station work and 5 minutes of running?
90 seconds sits just under the shortest station in a race. The two fastest are the sled push at a median 2.1 minutes and the sled pull at 2.6, measured in a peer-reviewed race simulation of 11 recreational athletes. This session goes shorter than the shortest because it repeats the handover far more often than a race does. The dose that matters here is how many times you run off a station.
The station work is prescribed in time rather than in reps or metres, so nothing in this session depends on a weight or a rep standard.
5 minutes of running is long enough that the disruption clears inside the rep. The problem the session is built around is the opening of a run. A rep that ends before your running settles trains the disruption without training the recovery from it. The runs are in minutes for the same reason the station work is. A distance invites a pace, and a duration gives runners of different speeds the same dose instead of the same ground.
Both durations are ours. The published compromised-run sessions use round numbers of metres and minutes and give no reason for any of them. These are picks with reasons behind them.
What pace or heart rate should the runs be?
Neither one. The run is governed by repeatability: the first rep sets the effort, and every rep after it matches. The check is ground covered, so the last 5 minutes should cover close to what the first did. Coaches who put a number on this work in distance reps and ask for the last rep within 5 to 10 seconds of the first.
A zone target won't do the job. Nearly all of a HYROX race sits inside the top two intensity bands, so a zone can't separate going too hard from going correctly. For why, see can you trust heart rate zones during a HYROX race? The session inherits the problem, because the run starts from whatever the station has already done to your heart rate.
Expect a high number and expect it to stay high, because a run started straight off a station is active recovery rather than rest. Why that is, and what the number does across a whole race, is covered in what does your heart rate actually do during a HYROX race? The station's own number runs high for a separate reason, which is that resistance work carries a heart-rate premium over running at the same oxygen cost. That premium is a fact about the station and says nothing about what you carry into the run.
So record heart rate and read it afterwards, against the last time you ran the session.
Which station should you use as the compromiser?
Burpee broad jumps by default. That pick is ours, for three reasons. It needs no equipment, so you can run the session this week wherever you run. It carries no load, so nothing here depends on a weight. And it's leg-dominant, which is the property the session needs.
Once you have a full station setup, the coaching sources agree on the swap: use the station that wrecks you most. Sandwich your weakest one, or ask which station is most likely to slow your next run down, and that's your compromiser.
Timing that choice has a framework, and it's RMR Training's rather than ours. Earlier in a block the compromiser can be a general movement, and it becomes race-specific as the race approaches. That comes from one coaching source rather than a documented consensus, and it ships here at that strength.
Whichever station you pick, keep it for the block. Changing the compromiser between sessions changes what the running is being compared against, and that comparison is the reason for holding it constant. The stations in race order are listed under what is the HYROX race format? Use whatever weight and rep standard your gym or your race division sets, because this session doses the compromiser in time.
How many rounds should you do, and how do you progress the session?
Start at four rounds and build to six. Four to six is where the independent coaching sources land, so the range is consensus. Starting at the bottom of it is our call, because the first time you run this the limiter is the handover rather than your endurance.
Progress by adding one round every two weeks, and stop at six. One published coaching progression states the rule in exactly that form. A second agrees on the axis, allowing a small increase in station work as the alternative. That's one documented source with corroboration on direction rather than a consensus, and it's worth no more than that.
The ceiling of six is ours. The source that states the progression carries it to eight; past six rounds this session costs the rest of your training week more than it returns. When six rounds are comfortable, change the compromiser or hold a higher standard on the running, and leave the round count where it is.
Everything else stays fixed while the round count moves. The work, the run, and the rest are what make one session comparable to the last. A week that changes two of them at once measures neither.
How often should you do compromised running?
Once a week. The coaching range is one to two sessions a week during the race-specific phase of a HYROX build. One is our pick inside it. This session is hard on the legs in a way that shows up over the next two days of running.
Put it on a hard day and keep the day after easy. It draws on the same resources as an interval session and a long run. Give those their own days rather than the ones either side of this. In a three-run week that usually leaves one easy run, one of these, and one longer run.
Two sessions a week is defensible close to a race, when the point of the block is race-specific fatigue rather than general fitness. It isn't where to start, and it doesn't hold up for long.
Assembling a full week and a full build is a bigger question than one session. A HYROX running plan is coming in this series.
How do you time the work and rest phases during the session?
By sound. The one boundary that has to be exact is the station into the run. It arrives with your hands on a sled, a ball, or the floor. That's the moment a wrist is unreadable. It's also where a few seconds of hesitation turns a compromised run into a rested one.
Set it up as a custom interval before you start: 90 seconds of work, 5 minutes of running, 90 seconds of rest. Repeat for however many rounds you're doing. RunBeat's Interval Training runs custom work and rest phases with a voice callout at every boundary. It also switches your Apple Music playlist between work and rest, so you hear the handover instead of looking for it.
Whatever you use, count the rounds. Losing a round or shortening a rest breaks the comparison against the last session, and that comparison is what tells you the training is working.
RunBeat isn't affiliated with or endorsed by HYROX. This session is ours, not a HYROX training product.
Sources
- Brandt T, Ebel C, Lebahn C, Schmidt A. Acute physiological responses and performance determinants in Hyrox© – a new running-focused high intensity functional fitness trend. Frontiers in Physiology. 2025;16. doi:10.3389/fphys.2025.1519240. 11 recreational athletes, simulated Individual Open race; values are medians. Source for the sled push and sled pull times and for the race intensity distribution cited above.
- Crisafulli A, Orrù V, Melis F, Tocco F, Concu A. Hemodynamics during active and passive recovery from a single bout of supramaximal exercise. European Journal of Applied Physiology. 2003;89(2):209–216. Source for the mechanism: heart rate stays up because muscular engagement keeps the drive to the heart up, not because metabolic demand stays up. 8 male amateur soccer players recovering from cycling, so the direction transfers and the size of the effect doesn't. No figure from it is stated here.
- Gotshalk LA, Berger RA, Kraemer WJ. Cardiovascular responses to a high-volume continuous circuit resistance training protocol. Journal of Strength and Conditioning Research. 2004;18(4):760–764. Source for one sentence only: resistance work carries a heart-rate premium over running at the same oxygen cost. It measured a circuit at 40% of one-rep max rather than a race station. No figure from it attaches to any station, and it says nothing about what carries into the run that follows.
- RMR Training: the practice of holding the compromising movement constant and reading the running against it, and the general-to-race-specific framework for choosing that movement. Paraphrased, and attributed to RMR Training rather than to an individual coach, because published spellings of the coach's name disagree.
- Coaching consensus across the published compromised-run sessions, read August 5, 2026: Hybrid Pro Coach, This Rox Life, Kracey, TWL, Xendurance, 26weeks, HYROX Vault, and Rox Lyfe. Source for the 4-to-6 round range, the zero-rest handover, the 60-to-120 second rest range, the repeatability standard, and the rule that the compromiser should be your weakest station. Where they disagree, or give a number without a reason, this page states its own pick and says so.
- Every duration, the round count, the frequency, the progression rate, the ceiling of six, and the default compromiser are RunBeat's. No coaching source or study prescribes this session as written.