Station by station, where a Hyrox race is actually won

Most athletes train the stations they enjoy and hope the rest holds up. The measured answer is blunter than that: four stations hold nearly all the gettable time, three are effectively settled, and the two biggest levers of all are not stations. Here is the whole race in order of what it will actually give you back, built from published race data with the honesty tags where the evidence thins.

Where the time actually is

Ranked by the gap between the 10th and 90th percentile across 12,479 results (Toulouse and Beijing, season 8):

A second dataset of 395,452 results reorders the top two but agrees on the shape: wall balls, sled pull, burpee broad jumps and sandbag lunges are where the race is won, and the two ergs plus the carry are nearly settled. All three of those combined hold less than a minute of realistic gain for most athletes.

But the largest levers aren't stations at all. Running fade costs the average athlete 1:53 between the first kilometre and the eighth. The roxzone separates recreational from elite by more than three minutes of standing still. Fix those before you buy a new sled harness.

→ Do this. Rank your training by that table, not by what you enjoy. If your row is 20 seconds off, leave it.

The runs: half your race, and it fades

Running is roughly 52% of a 1:25 finish, and it degrades in a pattern that gives the plan away. The average athlete runs 4:20 for the first kilometre and 6:13 for the eighth. The single biggest jump is from run one to run two, which is what going out too hard looks like in the data.

The fade scales with ability: a 1:20 athlete slows about a minute across the eight; a 1:40 athlete slows well over two. And the wheels come off in the same place for everyone: runs five and six show the sharpest step up, corroborated by a small lab study where run five was the longest of the whole race.

Jake Dearden puts the mechanism plainly: "If you go out too hot from the start, you're not going to recover and you're going to be chasing your tail for the entire event. Whereas, if you set off a little bit too easy, you can always push a bit harder."

There is also a physiological reason the runs feel worse than they should. In a peer-reviewed study of eleven recreational athletes, blood lactate was significantly higher after the stations than after the runs, meaning you begin every kilometre more acidotic than the running itself made you.

→ Do this. Hold within 15 to 20 seconds per kilometre from run one to run eight. More than 20 means you started too fast. Practise this against your own splits on the head-to-head view, where your run pace sits beside someone else's.

The roxzone: three free minutes

Transitions are the cheapest time on the course and almost nobody trains them. Measured across 825 athletes at one venue:

• Under 1:10 finish: 3:41 total roxzone, 28 seconds per transition
• 1:10–1:20: 4:33, 34 seconds
• 1:20–1:30: 5:19, 40 seconds
• 1:30 and over: 6:58, 52 seconds

That's 3:17 between the top and bottom groups, with no fitness component whatsoever. Because absolute times vary by venue layout, the transferable target is the percentage: keep transitions under about 6.5% of your total race time.

Two penalties live here. Entering the roxzone through the OUT arch, or leaving through the IN arch, is two minutes per infringement, and it's chip-detected, so there is no arguing it. Exiting a station through its entry arch is another two minutes. (The roxzone guide covers how to train the transitions.)

→ Do this. Jog every pass. Decide your walk-or-run policy before the gun, not at station four.

1. SkiErg: 1000m

The smallest spread of the eight, and that IS the finding: there is very little to win here, and a great deal to lose by taking it out too hard and paying for it on the two sleds that follow.

Medians: men 4:29, women 5:11.

Coach Jane Erbacher frames the trade exactly: "By putting in that 5% more effort you might only get off two to three seconds sooner, and it can cost you more later in the race."

The standards that catch people: the damper arrives preset at 6 and you may change it freely. Jumping technique is explicitly legal and your heels may overhang, but if a foot touches the floor rather than the base plate, that's an infringement. Leaving the SkiErg before 1000m is a disqualification, not a penalty.

→ Do this. Treat it as a controlled opener. Hip hinge, not arms.

2. Sled push: 50m

Medians: men 3:04, women 2:55, one of the few stations where the women's median is faster, on a lighter sled.

There's a caveat here worth more than any technique cue. Matched by finish-time bracket across four venues, sled push times swung 40 to 55 seconds between them at identical fitness levels. The dataset's own hedge is honest: likely the mat, the sleds, or something else not about the athletes. The rulebook now acknowledges the problem, reserving the right to declare an event "out of ranking" for championship qualification where venue conditions materially affect performance.

So don't pace this station off a target time. Pace it off feel on the day.

Weights, including the sled: women 102kg, women's pro and open men 152kg, men's pro 202kg. (Every division's weights.)

The standards: your assigned lane is mandatory, you never choose your own sled. The sled must fully pass the 12.5m mark before you turn. Missing a lane is a disqualification, detected by chip. And chalk is banned here; it's legal only at the sled pull and farmers carry.

→ Do this. Get low, arms locked, short fast steps, and never let it stop. Restarting a stalled sled costs far more than pushing slowly through.

3. Sled pull: 50m

One of the two biggest opportunities on the course, and the most rule-dense station in the race.

Medians: men 5:03, women 5:46. The most useful framing comes as a ratio rather than a raw number: elite athletes take only 18 to 30% longer on the pull than their own push, while average competitors take 74 to 78% longer. Compare your own two sled splits, that ratio tells you more than either time alone.

Jake Dearden, on his own channel: "Perfect your technique, I personally like to walk back with the rope. Keep the rope from under your feet so you don't trip. Stay relaxed, focus on your breathing."

Coach Paul Gillingham adds the one that's actually about the rules: "Use the box space to your advantage. Be careful of your feet at the back and the front, but if you are overly cautious, you will cost yourself time and energy by not fully utilising all the space you do have."

The standards: you may not step on the solid line at the front or back of the box while holding the rope, that's the classic sled-pull penalty, and it's a foot fault, not a sled fault. You must stay standing; no pulling from a seated or kneeling position. Warning, then 15 seconds per infringement.

→ Do this. Compare your pull against your own push. If the ratio is near 75%, this is your station.

4. Burpee broad jumps: 80m

Second-widest spread in the race, and the leak is rep count rather than rep speed. At 2.0m per jump that's about 40 reps; at 1.3m it's about 62. Every 10cm of jump distance removes two to four burpees.

Medians: men 5:22, women 6:24.

Megan Jacoby's cue is the one most amateurs get backwards: "Use the step-up method (stepping up one leg at a time after performing the burpee) to conserve energy and keep your heart rate lower. Maintaining steady movement with less breaks will yield a better time overall on this station than going too fast and having to take lots of pauses." And the step-up is explicitly legal, the rulebook permits using a knee coming out of the bottom.

Four standards that published guides routinely get wrong:

• The chest standard is the nipple line making clear contact with the ground. "Chest and thighs" is a CrossFit standard, not a Hyrox one.
• Both feet take off and land together, with a 5cm stagger tolerance, and no extra steps or shuffles at any point.
• Hands go no more than 30cm forward of the toes of your front foot, and once placed they cannot move forward.
• If either foot lands on the finish line, you do another burpee broad jump.

Lane lines here are guidance only, so overtaking is allowed.

→ Do this. Longer jumps, fewer reps, steady rhythm. Measure your jump in training and count what 80m actually costs you.

5. Row: 1000m

The narrowest spread on the course. This is not where your race is won, and the real cost of a hot row lands on the farmers carry and the run that follow it.

Medians: men 4:49, women 5:25.

Paul Gillingham: "If you reduce your pace a touch before the end of the station, you will cost yourself a few seconds at the most on the row, but can go into the subsequent run feeling much fresher."

The standards: damper preset to 6, footplates to position 4, both adjustable. Grabbing the handle before both feet are on the footplates is an infringement, a warning the first time and 15 seconds every time after, easy to trip in a rushed transition off the burpees. Resting mid-row is legal and unlimited. Stand up or leave the seat before the monitor reads 1000m and it's a disqualification (incomplete station); stay seated until the judge confirms.

→ Do this. Settle in, and ease the last 200m.

6. Farmers carry: 200m

The narrowest opportunity of the eight. The real time here is in the transitions either side, not the carry.

Medians: men 2:09, women 2:17. Weights: women 2×16kg, women's pro and open men 2×24kg, men's pro 2×32kg.

Jake Dearden: "Keep the kettlebells as still as you can. If the kettlebells are bouncing up and down in your hands, it is going to fatigue your shoulders. Try to relax. Long straight arms and a relaxed grip on the kettlebells is super important for efficiency. Move fast! If I told you to just stand with a pair of kettlebells in your hands for 3 minutes, your muscles would start to burn."

The most mis-stated rule in amateur coverage: putting the kettlebells down to rest is permitted, provided they don't move forward as you place them. The only cost is the clock.

And a free save almost nobody knows: failing to return the bells to the right box, or leaving them handles-down, is a 30-second penalty: waived entirely if you go back and fix it before you exit the roxzone.

→ Do this. Chalk up, relax the grip, move fast. And check the bells before you leave.

7. Sandbag lunges: 100m

The most counter-intuitive finding in this whole piece. Athletes go conservative here to "save the legs" for the final run. Across 134 athletes inside a tight 1:27–1:33 finish bracket, the correlation between lunge split and run eight was r = −0.08 for men and −0.07 for women, statistically nothing.

In the analysts' own words: athletes who crushed the lunges in 3:30 ran similar final kilometres to those who took 7:00. The trade you think you're making doesn't exist.

Medians: men 5:13, women 5:07. Weights: women 10kg, women's pro and open men 20kg, men's pro 30kg.

Megan Jacoby: "A slight lean forward in the lunge will place more emphasis on your glutes, which is, of course, your biggest muscle. When it comes to pacing, stepping through your lunges is the fastest way to go, but taking a small pause at the top of the lunge movement is the best way to conserve energy."

The standards: the trailing knee must clearly touch the ground, every rep ends standing tall with hips and knees extended, and lunges must alternate. No steps or shuffles between reps. Resting the bag on the ground, the barricades or your feet is 15 seconds per infringement, not a disqualification, whatever you've read.

→ Do this. Go harder than feels wise. The data says the last run doesn't care.

8. Wall balls: 100 reps

The largest single time sink in the race and the widest spread of any station: an average of 7:28, with the 10th-to-90th range running from 3:49 to 9:04.

Medians: men 6:52, women 6:30. Weights and targets: women 4kg to 2.70m, women's pro and open men 6kg (2.70m and 3.00m respectively), men's pro 9kg to 3.00m.

Cole Learn, on why the plan matters more than the effort: "I know for myself, if I set the ball down, I'm toast." Elli Stenfors offers the counterweight: "I tell myself one more rep, and then I can break if I need to." And a useful bit of perspective from Jake Dearden, he says he has still never gone unbroken on wall balls in a race.

The standards, where most guides are wrong:

• No warnings and no time penalty. It's a rep or a no-rep, full stop. The only time penalty at this station is two minutes for powdered chalk.
• Depth is hips below knees, in the rulebook's own words.
• You must start standing tall, you cannot pick the ball off the ground and immediately throw.
• You may catch the ball or let it drop, but you may not catch it after a bounce and continue.
• The digital screen is not the authority. The physical target and the judge are.
• Leaving before 100 valid reps is a disqualification.

(The full wall ball guide breaks the station down rep by rep.)

→ Do this. Plan your sets before you pick the ball up, and guard your depth as fatigue builds. A rep at 90% depth costs the same effort and scores zero.

Where these numbers come from

The movement standards, weights and penalties are taken directly from the HYROX Singles Rulebook, season 26/27, the book that governs Toronto in October. The split medians and percentile spreads come from published analyses of large public result sets (395,452 results across seasons 7–8; 12,479 from two season-8 races; 825 athletes for the roxzone; 134 for the lunge correlation), which are independent analyses of official results rather than official data themselves. The elite quotes are from named interviews and the athletes' own channels.

Two honest caveats. The widely-circulated "elite versus average" comparison table that most Hyrox blogs quote dates from 2022, uses about 15 athletes per side, and its author openly flags his own methodology, we've used it for the direction of a finding, never as a benchmark. And the venue-to-venue sled variance is a real measured pattern with no published methodology behind the matching, so treat the exact seconds as indicative.

You know where your time is. Now find out how far off it actually is.

See your splits beside another athlete's