A HYROX simulation is the race rehearsed in your own gym: eight 1km runs, each followed by one of the eight stations at your division’s load, in race order, on one clock and with no rest. The full version is the whole race. The half version is four runs and four stations. A station-only version strips the running out so you can see what the running costs you.
Most people treat a simulation as a test they either pass or fail. That wastes it. Done well, a simulation is a measurement, and its 16 splits tell you exactly where your race will go wrong if you change nothing: the run you went out too fast on, the station where the wheels came off, the minutes the running is stealing from your wall balls.
This page is the session written out, the way we programme it. You get the full simulation, two half versions and a station-only control, what to time, and a four-step way to read the result. Every distance, rep count and load comes from the official HYROX Season 26/27 Singles rulebook , published on HYROX’s rulebook page .
Key Takeaways
- A full HYROX simulation is 8 x 1km runs, each followed by one station in race order: SkiErg, sled push, sled pull, burpee broad jumps, rowing, farmers carry, sandbag lunges, wall balls.
- Use your division’s Season 26/27 loads. Open women push 102kg and Open men 152kg, both including the sled, and Men Open and Women Pro share the same loads.
- A half simulation keeps the stations at full race distance and halves the race instead: Runs 1 to 4 with the first four stations, or Runs 5 to 8 with the last four.
- A station-only simulation, all eight stations with rest instead of runs, is the control that shows how much the running costs each station.
- Press the lap button 16 times and you have the whole result. The run and station splits must add up to your total before you read anything else.
- Read the gap between Run 1 and Run 8 first. If Run 8 is more than 15 to 20 seconds per kilometre slower than Run 1, Run 1 was almost always too fast, not your fitness too low.
- Two or three full simulations per training block is enough, and the last one belongs at least three weeks before race day.
What a HYROX Simulation Is, and What It Is For
A HYROX simulation is a training session that reproduces the race format closely enough that its splits mean something on race day. It has three jobs: to rehearse your pacing plan under fatigue, to show which station breaks down when it arrives after the runs that precede it, and to give you a repeatable benchmark you can compare across a training block.
It is not your weekly compromised session. A compromised session pairs a few runs with a few stations to train the adaptation, and it fits in an hour. The simulation is the longest and most expensive session in a block, which is why you do very few of them.
If you want the time budget for every session type, including how long a simulation takes compared with everything else, the guide to how long a HYROX training session should be covers it. This page is about what goes inside the simulation and what to do with the numbers.
The format is worth respecting because it is mostly running. A 2025 study in Frontiers in Physiology took 11 recreational athletes through a simulated HYROX and found a median time of 86.5 minutes: 51.2 minutes running and 32.8 across the eight stations (each part is its own median, so the two do not add up to the total). The stations, though, were where heart rate and lactate peaked. That is the tension a simulation exposes: the running decides your time, and the stations decide how well you can run.
The Full HYROX Simulation, Written Out
The full simulation is the race in order: run 1km, do Station 1, run 1km, do Station 2, and so on until 100 wall balls, with the clock running throughout. Warm up for 10 to 15 minutes first with easy running, a few strides and a light set on the SkiErg and sled.
The Session
This table is the whole session in race order, with the load for each of the four singles divisions from the Season 26/27 rulebook. Sled loads include the sled.
| Round | Run | Station | Women | Men / Women Pro | Men Pro |
|---|---|---|---|---|---|
| 1 | 1km | SkiErg 1,000m | Bodyweight | Bodyweight | Bodyweight |
| 2 | 1km | Sled push 4 x 12.5m | 102kg | 152kg | 202kg |
| 3 | 1km | Sled pull 4 x 12.5m | 78kg | 103kg | 153kg |
| 4 | 1km | Burpee broad jumps 80m | Bodyweight | Bodyweight | Bodyweight |
| 5 | 1km | Rowing 1,000m | Bodyweight | Bodyweight | Bodyweight |
| 6 | 1km | Farmers carry 200m | 2 x 16kg | 2 x 24kg | 2 x 32kg |
| 7 | 1km | Sandbag lunges 100m | 10kg | 20kg | 30kg |
| 8 | 1km | Wall balls 100 reps | 4kg | 6kg | 9kg |
The running total is 8km in every division, and the only thing the divisions change is the load. Notice that Men Open and Women Pro race identical weights, so a mixed training pair in those two divisions can share every implement, with only the wall ball target set at different heights.
Two rulebook details are worth copying into your gym. The wall ball target sits at 2.70m for women and Women Pro and at 3.00m for men and Men Pro, so mark the right height on the wall before you start. And the SkiErg and rower damper is preset to 6 for every division on race day, which you are allowed to change, so set it to 6 in training too unless you already know you race on a different setting.
How to Set It Up in a Normal Gym
A gym is not a race venue, and the set-up decides whether your simulations compare with each other. Three rules cover most of it.
- Measure the runs the same way every time. A 1km outdoor loop or a treadmill both work. Swapping between them from one simulation to the next does not, because the splits stop being comparable.
- Keep the sled at 50m, however short your lane is. If your turf strip is 10m, do five lengths per push instead of four. The total distance is what counts.
- Substitute only what you have to, and write it down. No SkiErg means a 1,000m row in its place, and no sled means the closest leg-drive effort you can repeat identically. Both are fine for training. Neither split compares with race day, so mark it in your notes.
Gym floors also differ from the race surface, so a sled at the correct load can feel lighter or heavier than it will in the arena. Treat your sled splits as a trend across simulations rather than a race-day prediction.
How Hard to Go
Run the plan, not a test. On your first full simulation of a block, hold your goal race pace on every run and treat every station as race effort you could keep going at. If you do not have a goal pace yet, the HYROX pace calculator turns a goal finish time into a target 1km split.
The reason is that a simulation run flat out tells you only that you are tired. A simulation run to a plan tells you whether the plan survives contact with the stations, and that is the question race day asks. Save a genuinely raced simulation for later in the block, once you have a plan worth testing at full effort.
The Half HYROX Simulation
A half HYROX simulation is four 1km runs and four stations at full race distance and full race load, back to back, on one clock. It takes 40 to 55 minutes and costs far less recovery than the full version, which makes it the simulation you will actually do most often.
Front Half or Back Half
There are two halves, and they test different things.
- Front half: Runs 1 to 4 with the SkiErg, sled push, sled pull and burpee broad jumps. This is the heavy half. It tells you whether your opening pace is sustainable and what the two sleds do to your legs on Runs 3 and 4.
- Back half: Runs 5 to 8 with rowing, farmers carry, sandbag lunges and wall balls. This is where races fall apart, because grip, lunge turnover and wall ball cadence are the things that degrade late. To make it honest, run it after a hard 2km at race pace, or after 20 minutes of steady running, so you arrive at Run 5 already working.
Alternate them. A front half in one fortnight and a back half in the next covers the whole race over a month without a full simulation’s recovery bill.
Why Not Halve the Distances
Some programmes halve everything instead: 500m runs, 500m on the SkiErg, 50 wall balls. It is a reasonable workout, but it is no longer a simulation, because none of its splits mean anything on race day. A 50-rep wall ball split cannot be compared with a 100-rep one, and a 500m run does not build the fatigue a 1km run does.
Keeping the stations at full distance means every split in a half simulation is a real race split you can compare with your last race, your last simulation, or your target.
The Station-Only Simulation
A station-only simulation is all eight stations in race order at race load, with 3 minutes of easy walking or jogging between them instead of the 1km runs. Time each station on its own. It takes about an hour and needs no running beyond the warm-up.
It has two uses. The first is as a control: these are your station times on relatively fresh legs, and comparing them with the same stations inside a full or half simulation shows exactly what the running costs you, station by station. The second is practical. When a calf or Achilles niggle has stopped your running for a week or two, the station-only version keeps your station work at race specificity without the impact.
Run it near the start of a block to set a baseline, and again within a week of a full simulation so the comparison is fresh.
What to Measure in a HYROX Simulation
Measure 16 splits and the total, and the rest is optional. Press your watch’s lap button each time you finish a run and each time you finish a station, so the walk from the run to the station folds into the station split. That gives you 16 laps, which is the same structure a HYROX result uses.
This table is what to record and what each measurement is for.
| Measurement | How | What it tells you |
|---|---|---|
| Total time | Watch or gym clock | Your benchmark across the block |
| 8 run splits | Lap button at the end of each run | Pacing and fade |
| 8 station splits | Lap button at the end of each station | Your weakest station |
| Heart rate per run lap | Chest strap or watch | Whether a slower run was harder or just slower |
| Substitutions | A note | Which splits not to compare |
| Wall ball sets | Count them, for example 25, 20, 15 | Where the cadence broke |
Notice what is not in the table: your average heart rate for the whole session. It hides everything that matters, because a session whose first half was too easy and whose second half fell apart can average out to a perfectly reasonable number.
How to Read Your HYROX Simulation Result
Read a simulation in four steps, in this order: check that the splits add up, compare Run 1 with Run 8, find the station that fell apart, then work out the compromise cost from your station-only times.
Step 1: Check the Sum
Add your 16 splits and check they match the total. If they do not, a lap was missed or pressed twice, and every conclusion you draw from that point on is built on a wrong number. It sounds pedantic. It is the same rule your race targets have to pass: a set of splits that does not add up to the finish time is not a plan.
Step 2: Compare Run 1 with Run 8
Put your eight run splits in a row and look at the shape. Some fade is normal: a healthy simulation has its slowest run around Run 5 or Run 6, straight after the burpee broad jumps and the row, with Run 8 inside 15 to 20 seconds per kilometre of Run 1. If Run 1 is your quickest by a wide margin and the slide keeps going to Run 7 or Run 8, Run 1 was too fast, not your fitness too low. That fast start is the most common pacing failure we see, and the guide to what pace to run in HYROX shows the healthy curve and the failing ones side by side.
Check your heart rate alongside the splits. A late run that is slower at the same heart rate as Run 1 is fatigue showing up in the legs. A late run that is slower at a higher heart rate means the engine is the limit.
Step 3: Find the Station That Fell Apart
Look for the station that was furthest off your target, and then look at how it broke.
Wall balls that start in sets of 25 and end in sets of 5 are a cadence and leg problem. A farmers carry with three put-downs is grip. Lunges that slow every 10m are leg endurance. Each points to a different session in the next few weeks.
Step 4: Work Out Your Compromise Cost
Subtract each station’s station-only time from the same station inside the simulation. The difference is what running into that station costs you. A small gap means the station is strong under fatigue. A large one, especially late in the race, is where compromised training pays most, and our sample HYROX training weeks show how a compromised session is built.
A Worked Example
Take Priya, racing in the Women division eight weeks from now. Her first full simulation of the block, run in her gym with stations including the walk to them, comes out like this.
| Round | Run | Station |
|---|---|---|
| 1 | 5:10 | SkiErg 4:50 |
| 2 | 5:25 | Sled push 3:40 |
| 3 | 5:40 | Sled pull 4:50 |
| 4 | 5:35 | Burpee broad jumps 4:40 |
| 5 | 5:50 | Rowing 5:00 |
| 6 | 6:05 | Farmers carry 2:00 |
| 7 | 6:20 | Sandbag lunges 4:30 |
| 8 | 6:30 | Wall balls 6:30 |
| Total | 46:35 | 36:00 |
The sum checks out: 46:35 of running and 36:00 of stations make 1:22:35, which is what her watch shows. Her runs get slower every round bar one, and Run 8 at 6:30 is 80 seconds per kilometre slower than her 5:10 opener. That is the fast-start shape, far outside the 15 to 20 seconds of a healthy curve, so the first reading is that Run 1 was too fast.
Her station-only session the week before had wall balls at 5:00 and lunges at 3:50. So running into them cost her 1:30 on the wall balls and 0:40 on the lunges, and the wall balls are the station to target.
The fix is pacing. Eight runs averaging 5:50 add up to 46:40, five seconds more running than she did on paper, and every station after Run 1 gets legs that have not already been spent. So her next simulation opens at 5:45, and her next three weeks add wall balls straight off a hard run.
What the Result Does Not Tell You
A simulation time is not your race time. A gym has no Roxzone, the transition area between the run course and the stations that adds minutes to every race, and the guide to the Roxzone covers what it costs and how to cut it. Against that, a gym has no crowd, no marshalled lanes and no race-day adrenaline. Use the simulation to compare yourself with yourself, and to test a pacing plan, rather than to predict a finish time to the minute.
How Often to Run a HYROX Simulation, and When to Stop
Do two or three full simulations in a training block, three to four weeks apart, with the last one at least three weeks before race day. Do a half simulation every two to three weeks in between, and use the station-only version as a baseline at the start of the block and as the stand-in whenever your running is off.
The full simulation is expensive. Treat it as the hardest session of its week, with easy days after it rather than another hard session, and expect your legs to take most of a week to come back. That recovery cost is exactly why it does not belong in race week: the race week guide explains why a full simulation that close to the race leaves you arriving depleted rather than sharp.
A simple pattern for a 12-week block looks like this.
| Weeks | Simulation |
|---|---|
| 1 | Station-only baseline |
| 3 | Front half |
| 5 | Full simulation, run to plan |
| 7 | Back half |
| 9 | Full simulation, raced |
| 10 | Station-only or front half |
| 11 to 12 | No simulations, taper and race |
Weeks 11 and 12 are empty on purpose. The fitness from a simulation takes longer to arrive than the fatigue takes to leave, so a late one costs you more on race day than it can give back.
FAQ
A HYROX simulation is a training session that reproduces the race in your gym: eight 1km runs, each followed by one of the eight stations at your division's load, in race order and on one clock. A half simulation does four runs and four stations, and a station-only version does the eight stations with rest instead of runs.
A half simulation of four runs and four stations takes 40 to 55 minutes. A full simulation usually takes your goal race time plus 10 to 15 minutes, for moving between stations and loading the sleds yourself, which a race does for you. Add 10 to 15 minutes for the warm-up.
Two or three full simulations in a training block is enough, three to four weeks apart. In between, a half simulation every two to three weeks gives you most of the race-specific stress at a fraction of the recovery cost.
No later than three weeks before the race. A full simulation takes most of a week to recover from, and the fitness it builds takes longer than that to arrive, so a late one costs you more on race day than it gives back. Never do one in race week.
Use your division's loads from the Season 26/27 rulebook. Women push 102kg and pull 78kg on the sled, carry 2 x 16kg, lunge with 10kg and throw a 4kg wall ball. Men and Women Pro use 152kg, 103kg, 2 x 24kg, 20kg and 6kg. Men Pro use 202kg, 153kg, 2 x 32kg, 30kg and 9kg. Sled loads include the sled.
Yes. Swap in the closest leg-drive effort you can repeat the same way every time and write the substitution down. The session still trains you, but those two splits will not compare with race day or with a simulation done on a real sled.
No. A gym simulation has no Roxzone, which adds real minutes to every race, and it has different sleds and no race-day adrenaline. Use the simulation to compare yourself with your last simulation and to test a pacing plan, not to predict your finish to the minute.
Conclusion
A HYROX simulation is the race in your gym, and its value is in what you read from the splits afterwards. Run the full version two or three times a block, the half version every two to three weeks, and the station-only version as your control. Hold your goal pace instead of racing it, press the lap button 16 times, and read the result in four steps: check the sum, compare Run 1 with Run 8, find the station that broke, and work out what the running cost it.
Most athletes who do simulations stop at the total. The total is the least useful number on the sheet.
That reading is the part Kracey does every time. It is an AI HYROX coach that lives in Telegram, with nothing to install.
It reads the session from your watch through intervals.icu, judges it from your lap splits and heart rate rather than the whole-session average, and messages you with what it actually was. When it names target splits for your next simulation or your race, they add up to the finish time, because that check is built into the coach. The first 14 days are free, starting when it can see your training rather than the moment you sign up.
Meet your coach in Telegram and get your next simulation read for you .