Tapering for HYROX: what to cut and what to keep

The best-supported taper is about two weeks long, with training volume cut by 41 to 60% and the intensity and frequency kept the same. That's from a meta-analysis of 27 studies of competitive athletes.¹ For HYROX, that means fewer kilometres and fewer sets in the last fortnight, but still some running at race pace and still some heavy lifting.

Nobody has tested a taper for HYROX itself. Everything below is the general taper research applied to a race with running and strength in it, plus one real race week.

What the taper research found#

Bosquet and colleagues pooled 27 studies of competitive athletes where performance was measured in a real competition or a field test.¹ The best results came from:

• a taper of about two weeks (effect size 0.59);
• volume cut by 41 to 60%, reduced progressively rather than all at once (0.72);
• intensity kept the same (0.33);
• training frequency kept the same (0.35).

Mujika and Padilla's review lands in the same place with wider ranges. Keep the intensity, cut volume by up to 60 to 90%, and cut the number of sessions by no more than about 20%. Tapers in the studies they reviewed ran from 4 days to more than four weeks, and progressive tapers did better than a sudden drop.² They put the usual improvement at about 3%, with a range of 0.5 to 6%. That's a typical figure across the studies they reviewed, not a number you can bank.

The one thing both agree on is intensity. Mujika's later review is blunt about it: you can cut a lot of volume without losing performance, but the load shouldn't come down at the expense of intensity.³ A taper made of only easy jogging is a detraining block, not a taper.

Two qualities, one taper#

The worry with HYROX is that you have two things to protect. If you cut your lifting for two weeks, does the strength go?

It doesn't, as long as the weight stays heavy. Spiering's review found strength and muscle size held for up to 32 weeks on one session a week with one set per exercise, as long as the load stayed where it had been. Endurance held for up to 15 weeks on two sessions a week, or a third to two thirds less volume, as long as the intensity stayed.⁴ Two weeks is well inside both.

For lifting, that translates into very little. One or two sessions in the last fortnight, with one or two sets of your main lifts at your usual working weight, keeps the load on the bar familiar with very little volume. Put the last one early in race week rather than two days out.

So the risk in a HYROX taper isn't losing fitness. It's arriving tired, or arriving flat because everything went easy. The way through is the same for both halves: keep the sessions short and keep a little bit of each at full intensity.

A HYROX taper in practice#

Here's how that translates. It's our application of the research, not a tested protocol.

The volume steps add up to the 41 to 60% the meta-analysis found, spread over two weeks rather than dropped all at once.¹ Keep the frequency close to normal.² A cut of no more than about 20% means that if you usually train six days a week, race week is five days, not three. The sessions get shorter. They don't disappear. And keep the intensity in both halves: race pace on the run, normal working weights on the bar.³

Then there's station work. Sleds, lunges and 100 wall balls are heavy, high-volume work, and in a taper they behave like volume. A short brick two weeks out keeps the movement familiar. In race week, there's nothing left to gain from a hard station session.

What my race week looked like#

I raced HYROX Bangkok on Friday 14 August 2026, starting just before 9 in the morning, and finished in 1:13:38.⁵ Here's the running from my training log for the five weeks before it, in seven-day blocks ending the day before the race.

Ken's weekly running before HYROX Bangkok

	Value (km)
10 to 16 Jul	25.6
17 to 23 Jul	24.3
24 to 30 Jul	20.1
31 Jul to 6 Aug	20.1
7 to 13 Aug	10.8

Device-recorded runs from Ken's hybridAF log (Garmin and Strava), seven-day blocks ending 13 August 2026, the day before the race. Read-only query of 2026-10-26.

The last two full weeks were the same, 20.1 km each, and then the race week dropped to 10.8 km. Indoor riding, which was a big part of my training, went from 325 and 310 minutes in those two weeks to 123 minutes in race week.

Day by day, race week went like this:

The last long run was on 2 August, 15.06 km, and the last HYROX circuit was on 4 August, three rounds of a 1 km run, 30 wall balls and 22 m of burpee broad jumps. That's 12 and 10 days out.

Against the research, the volume cut was in the right range. Running fell 46% and riding about 60% compared with the two weeks before. Intensity stayed: the 2.73 km run on Monday averaged a heart rate of 174, higher than my 158 average across the race itself. But it was a one-week step, not a two-week progressive taper. The week before race week wasn't reduced at all.

I finished under my sub-1:15 target, with my runs ranked 127th of 3,254 men.⁵ That's one race, though, and it says nothing about whether a proper two-week taper would have been faster. If you have the choice, the meta-analysis points to starting the cut a week earlier than I did.

The last two days#

The research is about the final weeks, not the last 48 hours. For the day before, the useful rule is simply not to add fatigue: an easy 20-minute jog or nothing at all. Mine was a 22-minute run at an average heart rate of 139.

On race morning, warm up as you would for a hard training run. The first kilometre is run 1 of 8, not a sprint. In my race, run 1 took 4:35, slower than every other run except the last one into the wall balls.⁵

What hybridAF does in the last two weeks#

If you set a race with a date in hybridAF, the weekly plan changes as it gets close. From 14 days out it's in a taper phase: every suggested session is capped at 60 minutes, and only the shortest hard session stays hard. Any other hard sessions are eased to moderate, and hard runs are rebuilt as easy runs. In the last 3 days the cap drops to 45 minutes. The day before, any hard session is eased to moderate. On race day itself, the only thing it will suggest is an easy shakeout of 25 minutes or less. Those limits are fixed rules in the code, applied after the plan is generated, so they hold whatever else the plan says. It's the same idea as the research: shorter sessions, one bit of intensity kept. The planning your week guide covers how a race goal changes the plan.

Before that, from 15 to 35 days out, the app is in its sharpen phase, which asks for the most hard work of any phase. The switch to the taper lands two weeks out, which is where the research puts it.

Where this fits#

The taper is the end of a longer plan. Periodizing a hybrid year covers how the weeks before it are built, and deload weeks covers the lighter weeks during a block, which look similar but do a different job. For the race itself, my first HYROX breakdown goes through every split, and the HYROX hub has the rest of the series. hybridAF is not affiliated with HYROX.