He blocks out three weeks, clears the calendar, flies out to altitude, and grinds through the hardest training block of his year — the kind of camp riders talk about for a whole season afterward. He comes home and re-tests his numbers on the same climb he did before he left.

Nothing's moved.

Meanwhile a training partner who never left home — who spent the same three weeks sleeping in an altitude tent in his own bedroom — comes back faster. The mountain wasn't the variable that mattered.

Altitude training is sold as close to a guarantee: go high, come back with more red blood cells and a bigger engine. The pitch skips the part where a meaningful share of riders who do exactly what the camp asks come home with nothing to show for it — not because they cut corners, but because their own physiology never got the memo.

Section 01What the data actually says

Altitude is supposed to work through a fairly simple chain: lower oxygen availability triggers a rise in erythropoietin (EPO), EPO drives the bone marrow to make more red blood cells, more red blood cells means more oxygen-carrying capacity, and more oxygen-carrying capacity means a higher VO2max back at sea level. The "train low" half of live-high-train-low exists so the athlete can still hit real training intensities while that chain plays out, instead of just surviving thin air.

The foundational study here put 39 collegiate runners through 28 days living at 2,500m, training on one of three schedules — some training high the whole time, some training down at 1,200-1,400m, some mixing base-work high with intervals low — then sorted every runner by what actually happened to their sea-level 5,000m time. Seventeen improved by enough to count as a clear response. Fifteen either didn't improve at all or got slower. The seven runners in between — a small, ambiguous change either way — were set aside so the two groups stayed genuinely distinct (Chapman 1998).

Responders showed a significantly larger rise in erythropoietin (EPO) within the first 30 hours at altitude, and that EPO stayed elevated after two weeks; nonresponders' EPO had drifted back toward sea-level values by then (Chapman 1998). That EPO response is what drives the downstream chain — more red blood cells, more oxygen-carrying capacity, a real VO2max increase. Responders got all three. Nonresponders got none of them (Chapman 1998).

The wrinkle is that it isn't purely a biological coin flip either. Nonresponders in the same data also showed a significant slowing of their interval-training velocity at altitude, and pulled a smaller oxygen consumption out of those intervals than responders did (Chapman 1998). The paper's own read on why the camp fails for some athletes has two parts, not one: not enough EPO response to the altitude they lived at, and not enough training intensity held at the altitude they trained at (Chapman 1998). Response is part physiology you don't control and part protocol you do.

Section 02Why "the mountain" isn't the active ingredient

If most of what separates a camp that pays off from one that doesn't is EPO response plus whether training intensity actually held up — not the postcard — the next question is obvious: does it actually have to be a real mountain?

A crossover trial put 16 well-trained triathletes through two separate 18-day live-high-train-low blocks a year apart, training at the same low altitude both times — one camp with the "live high" hours spent in a real mountain's thin, hypobaric air, the other with those same hours spent in a normobaric altitude tent (Saugy et al. 2016). On a cycling ergometer, VO2max rose 4.9% after the tent block and 3.2% after the real-mountain block — increasing to the same extent in both (Saugy et al. 2016). Peak cycling power output rose 7.1% after the tent and 4.9% after the mountain, again with no significant difference between the two (Saugy et al. 2016). A separate 3-km running time trial improved in both conditions too, with no reliable difference between tent and mountain at any of the three points they re-tested it (Saugy et al. 2016).

Same dose, same physiology, different postcard. The mountain view was never the mechanism.

That reframes the whole decision. A camp isn't a place you travel to and hope. It's a hypoxic dose you either deliver correctly or don't — and geography is one of the least important variables in whether it lands.

Section 03What this actually means if you ride, not run

The foundational responder data comes from distance runners, not cyclists — the classification in that study was based on 5,000m running performance (Chapman 1998), and no published study has run the same retrospective-plus-prospective design on a cohort of road or gravel racers. What generalises cleanly is the mechanism — EPO response, red-cell production, VO2max — because that chain is about blood and bone marrow, not about running gait or pedalling technique. What doesn't automatically generalise is the exact responder split, or how much a given cycling-specific performance test would move.

The reasonable working assumption is that cyclists sit somewhere on the same spectrum runners do — plenty of true responders, a meaningful chunk of non-responders, and a second lever (holding real intensity at the "train low" altitude) that's about protocol, not biology. That's not a reason to skip a camp if you can afford the time. It's a reason to stop treating "I went to altitude" as interchangeable with "I got fitter," and start treating it as a hypothesis you check.

Section 04The protocol that actually respects the coin flip

1. Test before you book anything.

   Set a repeatable sea-level benchmark — a known climb, a time-trial course, a standard effort — before you commit the time to a camp. You need a real "before" to know if there was an "after."
2. Treat it as a dose question, not a destination question.

   The physiological trigger is altitude exposure hours at sufficient elevation, not geography. A well-run tent block produced VO2max and power gains statistically indistinguishable from a real mountain camp of the same length (Saugy et al. 2016).
3. Protect your interval velocity at the "train low" altitude.

   Nonresponders in the foundational data also lost real interval intensity during the camp itself (Chapman 1998). If the low-altitude training site isn't low enough for you personally to hold your normal intervals, part of the payoff is already gone before biology even enters it.
4. Re-test the same benchmark straight after, not "how you feel."

   The groups in the foundational data were sorted after the fact by what the retest actually showed, not by who felt more wrecked (Chapman 1998).
5. Budget it as a real coin flip, weighted by execution.

   Under half of an identical cohort got the full intended response (Chapman 1998). Plan the season so a non-response doesn't sink your build — don't schedule the camp as the one thing standing between you and your goal race.

Section 05What responder status doesn't tell you

None of this means a camp is worthless if the hematological chain doesn't fire. Altitude blocks still remove riders from normal life, tighten up sleep and eating discipline, and rack up genuine training hours in a way that can pay off through routes that have nothing to do with EPO — economy, pacing discipline, buffering capacity at altitude-simulated intensities. The specific claim this data supports is narrower than "altitude makes you fitter." It's "the hematological gains that camp marketing leans on are real for roughly half the field and absent for the rest, and hitting them takes both a real EPO response and holding real training intensity at altitude — not just showing up and suffering."

Section 06Applying it with THRIVE

Section 07Where riders get this wrong

Section 08Counterpoint: dose and sport matter

Responder status depends on getting a sufficient hypoxic dose — enough hours, high enough — so some "non-responses" in the wider literature are really under-dosing rather than a hard biological ceiling. The foundational cohort here were runners; the erythropoietic mechanism is generally assumed to generalise to cyclists, but the performance-transfer specifics haven't been tested head-to-head for riders in the same way. A camp can also still be worth running without a measurable red-cell response — economy and buffering-capacity gains may accrue on their own, separate from the EPO pathway this piece covers.

It's also worth being honest about what a coin-flip framing does and doesn't mean. It doesn't mean altitude training is a myth, or that the responders imagined their gains. It means the marketing around altitude camps assumes a guarantee the physiology doesn't provide, and that the only way to know which side you're on is to measure it — not to trust how the block felt.

Section 09Bottom line

Altitude camps aren't a guaranteed fitness cheat code. In the cleanest data available, under half the group got the intended physiological payoff — and the shortfall wasn't only bad luck; some of it tracked with nonresponders losing real training intensity at altitude, not just a blunter EPO response (Chapman 1998). The camp isn't broken and neither is every athlete who doesn't respond — but "I suffered enormously" was never the signal that mattered. The signal is the retest, and whether your intervals actually held up while you were up there.

If you're going to spend the time, treat it as a dose you can control rather than a place you have to fly to — a well-run tent can match the mountain (Saugy et al. 2016) — and test before and after so you actually know which side of the coin flip you landed on. Skip the test and you're left doing what most riders do after a camp: deciding whether it worked based on a feeling, in exactly the variable the data says can't tell you anything.