How to pace the eight HYROX runs

Pace the eight HYROX runs as evenly as you can, with run 1 deliberately controlled. That's what real finishers do: in Bangkok 2026, men around 1:00, 1:15 and 1:30 kept runs 2 to 7 within a few percent of each other, and run 7 was only about 2% slower than runs 2 to 4. The runs aren't recovery, either. Heart rate and lactate barely drop on them, so a pace that feels easy on run 2 has to still be holdable on run 7.

What the pacing research says#

There's no trial of HYROX pacing strategies. The closest thing is the general pacing literature. A 2008 review sorts pacing into a handful of patterns (negative, all-out, positive, even, parabolic, variable) and concludes that for events longer than about two minutes, a more even distribution of pace tends to give better times.¹ It also says that in the longest events, well-trained athletes usually drift into positive pacing, starting fastest and slowing. HYROX is an hour or more of mixed work, so it doesn't fit cleanly into any of the categories that review covers.

The one study with run-by-run data from a real HYROX race is on Pro men at HYROX Madrid 2025. Their average speed fell from 16.1 km/h on run 1 to 13.6 km/h on run 8, a progressive decline across the race. Averaged over the runs and the Roxzone, they moved at about 84% of their speed at the second ventilatory threshold and about 76% of the speed at which they hit VO2max in the lab.² So even these athletes weren't running anywhere near flat out, and they still didn't hold a flat pace. They faded. The question is how much fade is normal and how much is a mistake.

What real finishers actually do#

To get an answer that isn't a guess, we pulled the timing tables for every man who finished near 1:00, 1:15 and 1:30 at HYROX Bangkok 2026 and took the median of each run.³ Runs are measured mat to mat, from leaving the Roxzone to coming back into it. Run 8 is left out, because there's no Roxzone mat before the wall balls and its split includes the walk to the target.

Median run split by run number, Bangkok 2026 men

run	Around 1:00 (min)	Around 1:15 (min)	Around 1:30 (min)
1	4.03	4.67	5.37
2	3.43	4.03	4.67
3	3.58	4.25	5.03
4	3.57	4.18	4.97
5	3.63	4.3	5.1
6	3.57	4.25	5
7	3.58	4.25	5

Median of each run leg (Roxzone exit to Roxzone entry; run 1 from the start) for men finishing near 1:00, 1:15 and 1:30 at HYROX Bangkok 2026. Run 8 omitted because it includes the walk to the wall balls. Source: HYRESULT result pages, computed by hybridAF 2026-10-22.

Three things stand out, and they're the same at all three levels.

Runs 3 to 7 are nearly flat. After run 2, the medians sit within about 10 seconds of each other. Run 7 was a median 2.1%, 2.3% and 2.7% slower than the average of runs 2 to 4 in the three groups. That's much less fade than the Pro data suggests, partly because those numbers run all the way to run 8 and these stop at run 7.

Run 5 is the slowest middle run. It comes straight after the burpee broad jumps, and it was the slowest of runs 2 to 7 in every group. The lab simulation found the same: run 5 was its slowest run.⁴ If you have a run to protect with your pacing, it's the one after the burpees.

Run 2 is the fastest. It comes after the SkiErg and before the sleds, so most of the race's leg work is still ahead of you.

The run 1 problem#

In Bangkok, run 1 was the slowest of the first seven runs for every group: 4:02 against about 3:26 to 3:38 for the 1:00 men, 4:40 against 4:02 to 4:18 for the 1:15 men. Whether that's a crowded start wave, a longer first lap at that venue, or both, the timing can't tell you. The rulebook lets lap lengths vary by venue as long as the total is 8,000 m.⁵

Across all seasons, it's the other way round. HYRESULT's men's medians have run 1 as the fastest run (about 4:26), then 4:42 for run 2, about 5:06 to 5:17 for runs 3 to 7, and 5:45 for run 8.⁶ That's the positive pacing the review describes: start fastest, slow down.

So don't judge your race by run 1. If it's slow because of the start, you didn't lose anything you can get back by sprinting run 2. If it's fast, check it wasn't quicker than the pace you planned for runs 2 to 7.

Do you use the runs to recover?#

Physiologically, not really. In Brandt's lab simulation of a HYROX race, average heart rate on the runs was 169 bpm against 174 on the stations, and average lactate 6.2 against 6.6 mmol/L.⁴ Close to 80% of the race sat in the very hard heart rate band. The runs are a little easier than the stations, not easy.

What you can do is use the first 200 to 300 m of each run to let your breathing settle before you settle into race pace. Nobody has tested that in a study. It's simply a way to avoid starting every run in debt.

What the sleds cost the next run#

The biggest single step in every group is from run 2 to run 3, the run straight after the sled push. It was 9 seconds slower for the 1:00 men, 13 seconds for the 1:15 men and 22 seconds for the 1:30 men.³ So the slower the group, the more the push costs the run after it, and it never really comes back: runs 3 to 7 sit at that new level for the rest of the race.

Extra time on run 3 after the sled push

	Value (s)
Around 1:00	9
Around 1:15	13
Around 1:30	22

Median run 3 minus median run 2, men at HYROX Bangkok 2026 finishing near 1:00 (n = 20), 1:15 (n = 62) and 1:30 (n = 98). Source: HYRESULT result pages, computed by hybridAF 2026-10-22.

So plan for it. Don't set your target from how run 2 feels, because run 2 is the easiest run of the race for almost everyone. Set it from what you can hold after the sled, and start run 3 at that pace rather than trying to match run 2.

A pacing chart from real finishers#

Here's the same data turned into something you can pace off. Run 8 is the median of each finisher's total run time minus their first seven legs, so it includes the walk to the wall balls.

Medians don't add up exactly, so the last column is the median run total, not the sum of the others. And it's one venue. Use it for the shape: a controlled run 1, a flat middle, and a run 8 that's always the slowest because of where it ends.

Setting your pace#

Work out your target running total, divide by eight, and that's your average. Then shade it. On run 1, hold back: aim for your average or a touch slower, and don't chase anyone.

On runs 2 to 7, run your average. If run 2 feels easy, that's normal, because it was the fastest run in every group, so don't read it as a sign your target was too slow. Expect run 5 to be your slowest middle run, and don't panic when it is. On run 8, run what's left. It ends at the wall balls, so arriving able to throw matters more than the last 30 seconds.

If you need numbers to start from, runs 2 to 7 in Bangkok averaged about 3:34 per run for the 1:00 men, 4:13 for the 1:15 men and 4:58 for the 1:30 men. The split targets page has the full set for each level, stations included.

One rule to know: if you're running 4:00/km or faster, the rulebook requires you to use the inside Fast Lane.⁵ And you count your own laps. The lap screens are only an aid.

My eight runs#

For a single real example, mine from Bangkok, measured mat to mat like the groups above, were 4:35, 4:00, 3:55, 3:55, 4:05, 3:56, 3:55 and 4:54, with an official run total of 33:14 that ranked 127th of 3,254 men.⁶ HYRESULT's own run cards round a few of those a second slower. Runs 3 to 7 sit inside 10 seconds, with run 5 after the burpees the slowest of them, the same pattern as the groups above. Run 1 was slow, as it was for the groups above. Run 8 includes the walk to the wall balls.

The difference from the 1:15 group is the level, not the shape: my runs 3 to 7 were 13 to 20 seconds quicker than their medians. The compromised running guide lines each of my runs up with the station before it.

Practising it#

You learn pace by running it tired, so most of the work is in bricks: a station, then a 1 km run at your target pace, repeated. Two things make those sessions more useful.

First, hold the target pace on every rep, including the last one. If rep 4 is 15 seconds slower, the pace was too fast for that day.

Second, practise the hard ones. The run after the burpees was the slowest middle run in every group above, and the run after the sled push was the first big drop from run 2. Those are the runs to rehearse.

Then check it on race day. Paste your finish time, your eight run splits and your station total into the HYROX splits analyzer. It shows how much you faded from run 1 to run 8 and how much time went into the Roxzone, which is the other place runs quietly lose time. More on that in Roxzone explained, and the rest of the series is on the HYROX hub. hybridAF is not affiliated with HYROX.