Most HYROX programs are written backwards. They open with race-specific intensity because it feels relevant, then bolt aerobic work on wherever there's room left. The body doesn't adapt in that order, and neither does this plan.
Ten weeks, five phases, built in the sequence your physiology actually responds to: benchmark, base, build, peak, taper. Every pace, wattage target, and sled load in weeks 2 through 10 is calculated from data you generate yourself in week 1, not a generic standard pulled off the internet.
The race, in one pass
Every HYROX race worldwide follows the same structure regardless of division: eight 1km runs, each immediately followed by one station, in a fixed order, SkiErg, sled push, sled pull, burpee broad jumps, rowing, farmers carry, sandbag lunges, wall balls. Eight kilometres of running total. Only the loads change by division.
The station that decides more races than any other isn't the heaviest one. It's wall balls, station eight, arriving after 7km of running and seven other stations. A 2025 Bundeswehr Munich physiology study recorded its highest blood lactate readings there, higher than at the sled. That's the reason this plan has athletes training wall balls every single week, not just once near the end.
What actually predicts a fast finish
The same study is direct about which physical qualities separate a strong finish from an average one, and it isn't grip strength or one-rep max.
A big aerobic engine, built before anything else, is what the data points to. Total endurance training volume and VO2max predicted finish time more strongly than grip or maximal strength. That's why the base phase is three weeks of aerobic threshold work before any race-specific intensity shows up, and why compromised running, running while your legs are already fatigued from a station, gets a dedicated session every week from week one. Running economy measurably degrades in the kilometre after a fatiguing station. It's a trainable skill, not a fixed trait.
The five-phase structure
Nothing in weeks 2 to 10 is a fixed number. Every pace, wattage target, and sled load is relative to what you record in week 1, then reset again at the week 6 deload.
| Weeks | Phase | Purpose | Sled load* |
|---|---|---|---|
| 1 | Benchmark | Test and calculate every zone, pace, and load used for the rest of the plan | Test only |
| 2–5 | Base | Aerobic threshold volume, movement technique, first exposure to compromised running | 55–65% |
| 6 | Deload | Volume down ~40%, retest key benchmarks, technique only on sled | 50%, technique only |
| 7–8 | Build | Lactate threshold work, race-order station density increases | 70–80% |
| 9 | Peak | The one and only near-full simulation in the plan, ~85–90% intent | 90–100% |
| 10 | Taper | Volume down sharply, short and sharp only, arrive fresh | 50–60%, light touches |
*As a percentage of your target division's official race weight, not a self-tested max.
One rule holds across all ten weeks: day one is always genuinely easy, a real conversational-pace aerobic session, not a "controlled" version of something hard. That's the day that lets the other three sessions actually produce adaptation instead of just accumulating fatigue on top of fatigue.
One simulation, not ten
Week 9 contains the only near-maximal, full-distance simulation in the entire program, all eight stations, race order, full 1km runs between each, at roughly 85 to 90% intent. Everything in weeks 1 to 8 builds toward executing that one session well. Everything in week 10 protects it by tapering hard.
A full-effort simulation is a high-cost session. Run it every week and you build fatigue tolerance instead of fitness. Placed once, 8 to 9 days out from race day, it banks pacing discipline and clean transitions without leaving you flat on race morning.
Get the full 10-week guide
The complete plan, all 24 pages: race format and station-by-station demand, official division weights, a full terminology glossary, the benchmark and wattage testing protocol, and the entire week-by-week program. Pair it with the interactive calculator to turn your own benchmark numbers into every pace, wattage zone, and sled load in the plan automatically.
Frequently asked questions
How long is the Bedrock HYROX training plan?
Ten weeks, split into five phases: one week of benchmark testing, four weeks of aerobic base, one deload week, two build weeks, one peak week containing a single full-distance simulation, and one taper week.
How many days a week does the program require?
It's written for four sessions a week. A three-day condensed option is included for athletes who can't train four days, built around the same principle of never stacking two hard days back to back.
Do I need to know my one-rep max before starting?
No. Week 1 is a full benchmark and wattage week that establishes your own run pace zones, erg wattage zones, and starting strength loads. Every target in weeks 2 to 10 is calculated from that data, not a generic standard.
How is sled load calculated in the plan?
As a percentage of your target division's official HYROX race weight, not a self-tested max. That number is fixed and known in advance, which removes the guesswork and the injury risk of testing a heavy, unfamiliar sled load in week one.
Why is there only one full simulation in the whole plan?
A full or near-full effort simulation is a high-cost session. Used as a recurring weekly test it builds fatigue tolerance instead of fitness. Placed once, roughly 8 to 9 days out from race day, it does its job, banking pacing and transition skill, without digging a hole you can't recover from.
Every number in this plan started as a guess until it was tested in week one. Trust your own watts, splits, and sled times over any fixed number on a page, including this one.