TRVL Train™
Race Science

Your Altitude Camp Isn’t Changing Your Blood. It’s Buying You a Training Block.

The one- and two-week camps amateurs actually do sit at or below the noise floor of the test used to measure them. The good news is that the thing they do deliver is worth more to most riders than the thing they don't.

Race Science · 6 min read
EXPECTED tHb-MASS GAIN · ~1% PER 100 h ABOVE 2,100 m MEASUREMENT ERROR <2% left of this line = indistinguishable from noise WHAT AMATEURS ACTUALLY BOOK 1 week · 168 h ~1.7% below the noise floor 2 weeks · 336 h ~3.4% point estimate — but the range includes zero WHAT THE RESEARCH MEASURED AT 3 WEEKS 3 weeks · 504 h 0% to +6% same mountain, same dose — hypoxic dose explains only ~17% of the variance AND THE DOSE ASSUMES YOU LIVE HIGH Sleep at 2,200 m 168 h/wk Sleep at 1,200 m, ride high 4 h/day ~28 h/wk a sixth of the assumed dose
Expected haemoglobin gain by camp length against the sub-2% measurement error, the observed spread at three weeks, and the effect of sleeping low.

Every spring, thousands of amateur cyclists book a week or two in the mountains. Most of them believe, somewhere in the back of their mind, that they are going to come home with better blood.

They are not. But they are going to come home fitter, and the reason why is worth understanding — because it changes how you should plan the trip.

The arithmetic, which is unusually clean

The published dose-response for altitude exposure is about 1% increase in total haemoglobin mass for every 100 hours spent above 2,100 metres. Live at altitude around the clock and you bank roughly 24 hours a day.

A one-week camp is about 168 hours. That predicts a gain of roughly 1.7%.

Now the number that changes how you read it. The standard method for measuring total haemoglobin mass — carbon monoxide rebreathing — has a measurement error below 2%.

The expected gain from a one-week camp is smaller than the error bar on the instrument used to detect it. Test your blood before and after a week at altitude in a good lab and the result would be indistinguishable from noise. Not small. Unmeasurable.

Two weeks is a coin flip, not a guarantee

Two weeks doubles the dose to about 336 hours, and the point estimate — roughly 3.4% — does clear the measurement error. So the honest answer for a fortnight is more interesting than a flat no.

It is this: the average says you gained something, and the spread says you might not have.

A 2026 review in Frontiers in Physiology pooled the past decade of altitude studies and found that hypoxic dose explains only about 17% of the variance in haematological response. The same review notes that three-week camps above 2,000 metres — a bigger dose than any amateur trip — produced haemoglobin changes ranging from zero to plus six percent. Same mountain. Same duration. Some athletes got everything the model promised and some got nothing.

At two weeks, with a smaller dose and that much variance, the distribution of plausible outcomes comfortably includes zero. A two-week camp is a coin flip on the blood. It is a certainty on the training load.

You are probably not even getting the dose you think

Here is the part almost nobody mentions, and it hits amateur camps hardest.

The model counts hours above the altitude threshold. Not hours in the mountains — hours above roughly 2,100 metres.

Most amateur camps are based in a valley town. Alpine and Pyrenean camp bases typically sit somewhere between 800 and 1,400 metres. You sleep low and ride up. If you sleep at 1,200 metres and spend four genuinely high hours a day on the bike, you are banking something like 28 qualifying hours a week — not 168.

That is roughly a sixth of the dose the arithmetic above assumed. The one-week camp that theoretically bought you 1.7% may have actually bought you about 0.3%.

The camps that work for blood are the ones where you live high. Very few amateur trips are built that way, because sleeping at 2,200 metres is inconvenient, the hotels are worse, and the riding is better lower down.

So what did you actually buy?

Something genuinely valuable, and probably more useful to you than 2% more haemoglobin would have been.

You bought a block of uninterrupted training. No commute, no work, no domestic logistics, no compressed weekday sessions. For most amateurs, the binding constraint on fitness is not oxygen-carrying capacity — it is that ordinary life caps them at eight or ten hours a week. A camp lifts that ceiling for a fortnight.

You bought terrain you cannot get at home. Sustained climbs of forty minutes to two hours produce a different stimulus from anything available to a rider on rolling roads, however hard they push.

You bought the conditions for recovery between hard days — sleeping properly, eating properly, and doing nothing between rides except being a cyclist. That is what makes back-to-back big days survivable, and back-to-back big days are where the adaptation comes from.

And you bought the focus. Two weeks of thinking about nothing else has an effect that is difficult to quantify and obvious to anyone who has done it.

None of that is a consolation prize. That is the mechanism. The fitness people bring home from altitude camps is overwhelmingly the fitness they would have brought home from the same training block at sea level — plus the small, unreliable, probably-unmeasurable blood component.

How to plan the camp for what it actually is

If it is a training block, plan it like one.

Choose the venue for the riding, not the elevation. A camp base at 900 metres with world-class climbs beats a base at 2,000 metres with mediocre roads, because you are buying training, not hypoxia.

Build the week around recovery, not heroism. The camp's value comes from consecutive quality days. The classic amateur error is going too hard on days one and two and spending the rest of the trip surviving.

Expect to ride slower than at home and do not fight it. At altitude your sustainable power is down, and chasing home numbers on a climb at 2,000 metres is how a camp turns into a hole.

If you genuinely want the blood adaptation, the requirements are specific: three to four weeks minimum, living at 2,000–2,500 metres rather than visiting it, with iron status checked well beforehand — a deficient athlete cannot build red cells no matter how long the camp. And accept that you may still be among the athletes who get nothing, because the dose explains only a sixth of the outcome.

That is a serious commitment, and it is the right one for a professional. For everyone else, the mountains are still worth the trip. Just be clear about what you are buying.

A note on sources. This is a field with commercial actors — altitude tent and generator manufacturers, altitude-house hotels, and camp operators — all of whom benefit from the belief that short hypoxic exposures reliably work. The sources here are deliberately independent of that industry and lean on researchers who have publicly argued the sceptical case.

Key takeaways

  • The dose-response is roughly 1% haemoglobin mass per 100 hours above 2,100 m. A one-week camp predicts about 1.7% — below the sub-2% measurement error of the test used to detect it.
  • Two weeks has a point estimate around 3.4%, but with dose explaining only ~17% of the variance and observed responses spanning 0 to +6%, the realistic range includes zero. A coin flip on the blood; a certainty on the training load.
  • Most amateur camps sleep low and ride high, so actual hours above the threshold may be a sixth of what the arithmetic assumes. Living high is what drives the adaptation.
  • What a short camp reliably delivers is a block of uninterrupted training, terrain unavailable at home, proper recovery between hard days, and total focus. That is the mechanism, not a consolation prize.
  • Plan accordingly: choose the venue for the riding rather than the elevation, structure the week around consecutive quality days, and expect lower power at altitude.
  • If the haematological adaptation is genuinely the goal: three to four weeks, living at 2,000–2,500 m, iron checked in advance — and accept you may still be a non-responder.

Sources