Train to Mountain builds a personalised plan around your specific peak and current fitness, then recalibrates it every Sunday against the training you actually did. It runs on the Banister fitness-fatigue model to time your form for summit day, polarised 80/20 intensity for the aerobic base, and peak-specific demand profiling across five training pillars: aerobic engine, muscular endurance, vertical capacity, eccentric descent strength, and an altitude strategy.
What "personalised" actually means here
Most mountaineering plans are templates: a fixed 16-week block everyone follows regardless of their peak or their starting point. That is a workout schedule, not a training plan. The problem is not the exercises, it is that the same block cannot be right for a fit hiker aiming at a 4000m (13,100 ft) glacier peak and a near-beginner aiming at the same summit next year. Personalised mountaineering training means the plan is shaped by three inputs at once: your objective, your current fitness and history, and the hours you can genuinely train each week. Change any one and the plan changes. For the fuller version of this argument, see why a workout log isn't a training plan.
The engine: the fitness-fatigue model
Underneath the plan is the Banister fitness-fatigue model (Banister et al., 1975). It treats your form on any given day as fitness minus fatigue. Every hard session builds both: it raises long-term fitness and adds short-term fatigue. Fatigue fades faster than fitness does, and that gap is the whole game. Modelling the two separately is what lets the plan load you through the season, then deliberately shed fatigue in a taper so your form peaks on summit day, not three weeks early. It is also why the honest answer to "am I ready" is not "how fit am I" but "what will my form be on the date I climb." More on the science behind this on the science page, and on the taper specifically in how to taper for a mountaineering climb.
The intensity mix: polarised 80/20
The aerobic base is the largest single lever in mountaineering fitness, and it is built mostly slow. Train to Mountain follows a polarised distribution: roughly 80% of training volume easy, in Zone 2, and about 20% genuinely hard, with little of the grey middle that feels productive but mostly accumulates fatigue (Seiler & Kjerland, 2006). This is the pattern that endurance research keeps returning to, and it maps cleanly onto mountain days, which are long and mostly aerobic with sharp intense sections. If Zone 2 is new to you, start with Zone 2 training for mountaineering and heart rate zones for mountaineering.
Profiled to your peak, not a generic mountain
A plan is only as good as the objective it is built for. A long non-technical glacier peak, a short sharp technical route, and a big high-altitude expedition demand very different balances of endurance, vertical capacity, descent strength, and altitude preparation. Train to Mountain reads the demand profile of your chosen peak (its vertical gain, typical summit-day duration, descent load, and altitude) and weights the training accordingly. You can see any peak's profile on its page in the Train for a Peak hub, and get a quick estimate of your current gap to it with the Summit Readiness Calculator.
The five training pillars
Every plan touches all five of these; the objective decides how heavily each is weighted.
Aerobic engine
The endurance base, built mostly at Zone 2. The biggest determinant of how a long mountain day feels.
Muscular endurance
Producing force for hours under a pack without fading. See the muscular endurance guide.
Vertical capacity
Specific uphill work that mirrors summit-day gain, on stairs, treadmill incline, or real hills.
Eccentric descent strength
The downhill loading that blows out under-trained quads (LaStayo et al., 2003). See eccentric training for descent.
Altitude strategy
How you prepare for and adapt to thin air (Bärtsch & Saltin, 2008). See altitude acclimatisation.
Recovery and load
Managing the fatigue side of the model so fitness sticks. See recovery and overtraining.
Why it recalibrates every week
Plans slip. You miss a week to work travel, catch a cold, or get a surprise big day in the hills. A rigid plan breaks at the first slip and then quietly stops matching your reality. Train to Mountain recalibrates once a week, on Sunday: it reads what you actually completed and adjusts the coming week's load and intensity. Weekly is deliberate, not real-time. It is frequent enough to absorb what life throws at the plan, but stable enough that you are not chasing a target that moves every hour. That single feedback loop is what keeps the fitness-fatigue model honest all the way to summit day.
Established science, applied to your peak and your week
Train to Mountain is not a new theory of training. It applies well-established endurance science (fitness-fatigue modelling, polarised intensity, eccentric and altitude-specific work) to your specific objective, your data, and the hours you actually have. Start with the Summit Readiness Calculator to see your gap, or read AI coach vs human coach for where an algorithmic coach is genuinely better and where it is not.
A good mountaineering plan is not the one with the cleverest workouts. It is the one that knows what mountain it is training you for, and adjusts when your week does not go to plan.
Common questions
What is personalised mountaineering training?
Personalised mountaineering training builds a plan around one specific objective and one specific athlete, rather than handing everyone the same template. At Train to Mountain it means the plan is shaped by your target peak's demands (vertical gain, summit-day duration, descent load, altitude), your current fitness and training history, and the weekly hours you can genuinely commit. The plan then recalibrates every week against the training you actually did, so it stays honest as life happens.
Does Train to Mountain use the Banister fitness-fatigue model?
Yes. The engine is built on the Banister fitness-fatigue model (Banister et al., 1975), which treats form as fitness minus fatigue. Every session adds both fitness and fatigue; fatigue fades faster than fitness. Modelling the two separately is what lets the plan build load through the season and then shed fatigue in a taper so your form peaks for summit day rather than too early or too late.
Is Train to Mountain an AI coach?
It is an algorithmic coach. The plan is generated and recalibrated by a training model rather than hand-written each week, which keeps it affordable and always current with your latest sessions. It applies established endurance science to your data, not guesswork. See our guide on an AI coach versus a human coach for where each is genuinely better.
How often does the training plan adapt?
Once a week, on Sunday. It reads the training you completed and adjusts the coming week's load and intensity. Weekly recalibration is frequent enough to absorb missed sessions, illness, or a big unplanned day, but stable enough that you are not chasing a plan that changes every hour. It is not real-time or daily re-planning.
Does the plan train me for my specific peak?
Yes. A long non-technical 4000m (13,100 ft) glacier peak, a technical alpine route, and a big high-altitude objective demand different balances of aerobic base, vertical capacity, descent strength, and altitude strategy. The plan weights the five pillars to your chosen peak. See a peak's demand profile in the Train for a Peak hub, and estimate your gap with the Summit Readiness Calculator.
What are the five training pillars?
Aerobic engine (the endurance base, built mostly at Zone 2), muscular endurance (force production for hours under a pack), vertical capacity (specific uphill work mirroring summit-day gain), eccentric descent strength (the downhill loading that blows out under-trained quads), and an altitude strategy (preparing for and adapting to thin air). Every plan touches all five; the objective decides how heavily each is weighted.