Downhill Running in Ultramarathons: The Cost of Free Speed

You get over the top of a climb and start taking time back right away. Breathing settles. Pace jumps. Runners who got away on the uphill start coming back to you. The descent is going well, so you press a little more.

An hour or two later the race feels different. The legs do not respond the same way. Foot placements take more attention. The next climb bites early. On technical ground you back off before you want to.

A lot of downhill mistakes happen like this. The descent feels manageable, so the runner keeps taking what it offers. Later the rest of the race pays for it. The bill can arrive as heavier braking, eccentric damage, lost technical control, or simply more descent load than the legs were ready to absorb.

It helps to understand what descents change in the muscles, joints, eyes, and nervous system - and to tell the difference between downhill discomfort that belongs to the work and downhill damage that starts shaping the rest of the day.

Who this is for

This article is for you if:

• you climb well enough but keep losing time, rhythm, or confidence on descents
• your breathing is still fine late in a race, but the legs stop cooperating on the way down
• you want to understand cadence, landing control, foot strike, and technical downhill control
• you want to train downhill running well enough to make it an advantage instead of something you simply survive

This article is not for you if:

• you treat downhill speed as a courage test instead of a skill and load problem

Why this matters

A fast descent can come with a price. The next climb starts costing more than it should. On the next runnable section the legs stop giving back as much. On technical ground you start backing off because foot placements take more attention and fewer of them land cleanly.

Downhill running is a major limiter in mountain ultras. It creates some of the biggest time gaps in races, and some of the biggest mechanical costs. Newer mountain runners often assume the race is decided uphill because climbing is the obvious work. In practice, many mountain runners discover that descents cost them more time than the climbs they trained so carefully for.

In long races, downhill running decides how much of your fitness you still get to use later. If you burn through the descents early, you carry that cost into the runnable sections, into the technical terrain, and into every climb that follows. Downhill speed can be borrowed early and paid for later.

The descent is not a break

Many runners still treat the descent like the easy part. Heart rate comes down a bit. Breathing settles. The climb is over. It is tempting to think this is where you get something back. On smooth terrain that can feel especially true.

In reality, the work has only changed shape. On descents you pay less through breathing and more through the legs, the joints, and how well the nervous system keeps up with the speed of foot placements and corrections. On smooth downhill the demand is mostly mechanical. On technical downhill it is mechanical and neural at the same time. You are absorbing impact, reading terrain, choosing lines, and correcting small mistakes quickly enough to keep moving well.

The descent is not a break. It is another demand, and in long races it can decide how much of your fitness is still usable later.

What the legs pay for

On flat ground and climbs, much of the running story is about producing force. On descents, it is about controlling force.

Every downhill step asks the body to catch a falling mass, absorb impact, and stop the lower body from collapsing into the next stride. The research term here is eccentric work: the muscle is producing force while lengthening. If you want a simple picture, think less of a spring and more of a shock absorber. On steep descents the knee and quadriceps take a large share of that job.

One treadmill study makes the point clearly. On a fairly steep downhill - roughly a -16% grade - the impact at each landing is about 54% higher than on flat ground, and the forces needed to catch and slow the body at each step are about 73% higher. That is a big jump in load for a section of the race that can feel easier because the breathing settles.

This helps explain one of the most common downhill surprises in ultra. The lungs feel fine, so the section seems safe. Meanwhile the quads are catching and slowing the body step after step. Over time that shows up as exercise-induced muscle damage, loss of force, slower response, and the feeling that the legs no longer catch you the way they did earlier. In training, the soreness and strength loss from a hard downhill bout can peak a day or two later and affect you for several days. In races, the same pattern can arrive earlier because the day keeps going. You are still climbing, still descending, and still adding load to tired legs. The result is less control, less stiffness where you want it, and less trust in the next foot placement.

The calves and plantar flexors are part of this too. A more forefoot-heavy descent can soften some of the impact peak, but it also shifts more work toward the lower leg. Downhill technique does not make load disappear. It redistributes it, ideally without moving too much onto a structure that is less ready for it.

What happens when you overdo the descents

The first price can be simple soreness. A hard downhill session or a race with serious descent load can leave you with strong DOMS for the next couple of days, especially if the eccentric work was new or bigger than what you had trained for. That is unpleasant, but manageable if you expected it and planned around it.

A more specific price can be lateral knee pain, patellofemoral irritation, overloaded calves, or an ankle that stops trusting technical ground. In a long race they can turn a fast descent into a section you can barely get through. In the worst case they stop the race altogether.

The chain is fairly simple. One structure gets tired first, so another one has to take more of the work. If the quads lose some of their ability to control the landing, the knee may start taking more of the stress. That can be one route into symptoms sometimes described as lateral knee pain or IT-band-related pain: too much downhill braking, too many repeated loaded steps, and not enough strength or control left in the system to keep the knee moving cleanly. If the lower leg is not ready for a more forefoot-heavy descent, the calves can start tightening and losing spring. If the ankles and feet stop reacting well on technical ground, every correction gets later and rougher. Bombing the descents too early, especially without the preparation to support it, may start as a pacing error and turn into a tissue-tolerance issue.

This distinction is important. General soreness is one thing. Sharp local pain, rapidly worsening control, or a structure that starts dictating every step is something else. That tends to mean the course, the strategy, or the preparation asked more of one system than it was ready to give. If pain is sharp, worsening, changes your stride, or does not settle, stop forcing the descent and get advice from a qualified medical professional.

Good downhill running starts with landing control

On descents you are always slowing the body a little at landing. In biomechanics this is often called braking. The practical question is how to spread that work out so no single step does more than it needs to.

Keep the stride shorter than your ego wants to

The classic downhill mistake is reaching too far in front. It feels fast because the ground is coming at you quickly, but it tends to mean harder braking on each landing, with a bigger hit to the quads and knees. Research from cadence studies on flat running, supported by downhill-specific work, points in a consistent direction: a modest increase in cadence, around 5-10% above your preferred rate, helps reduce how much slowing each step has to handle.

You do not need to chop your stride into something unnatural. The goal is to stop hunting for speed with your feet. On most descents, speed comes better from staying quick underneath yourself than from reaching farther downhill.

Match your body to the slope

Leaning back can feel safer, especially when the trail gets steeper. It can also make things rougher. Sitting behind your feet tends to mean more slowing at contact and can raise knee stress, especially around the patellofemoral joint.

A better starting point is a relaxed whole-body position that follows the slope: not diving forward, not sitting back, just staying over the feet with a slight forward angle that matches the hill. You want the body organised enough to keep moving, not braced against the descent.

Let foot strike adapt

There is no single perfect foot strike for downhill running. A more forefoot landing can reduce the size of the impact peak, but it tends to move more eccentric load to the calves. Landing more on the heel shifts more of that slowing work back toward the knee. On technical ground the best runners vary this naturally from step to step.

That is the more important part in practice. The trail changes under you. Your mechanics need enough range to change with it too.

How to read a descent from the body

One of the practical skills in downhill running is recognising the moment when you are still moving with the slope versus the moment when you are starting to fight it.

Early in a good descent the feet are landing where you expect them to. You do not need extra steps after each impact to tidy things up. The stride feels quick rather than forced. On technical ground you still feel like you are choosing lines instead of only surviving what is in front of you.

Then the picture changes. The foot starts landing a little farther ahead than you meant. The upper body gets tighter. On rocky terrain you stop looking ahead and start watching almost every step near your feet. On smoother descents the speed may still be there, but every landing starts to feel heavier.

That is a better warning sign than split pace. A lot of downhill trouble starts as a change in quality before it becomes a change in speed. The legs stop feeling responsive. Every landing feels a bit heavier. You stop moving cleanly over the trail and start correcting more from step to step.

This is where good descenders are different. They notice earlier when the downhill is starting to cost too much, and they change something before the section gets away from them.

Quick signals: what kind of downhill problem is this?

Not all descents ask for the same thing

A smooth fire-road descent and a rocky alpine chute are not the same problem. Once the terrain gets irregular, downhill running becomes as much about information and stability as about force production.

Runnable descents

On a smoother downhill, the main risk is not fear. It is greed. The trail invites a longer stride and a faster rhythm than the legs may be able to support later. This is where cadence, landing control, and posture matter most. If you are overstriding here, the cost can show up later as quads that no longer want to absorb the trail.

Practically, this is the descent where you want to feel quick rather than stretched out. Let the steps stay light, keep the knees soft enough to absorb instead of block, and resist the urge to prove how much speed you can hold early. On these descents the trail itself may be manageable. The trap is that it feels easy enough to invite a much bigger bill later.

Steep descents

This is where soft knees matter. Not floppy, just ready to absorb. If the legs are locked and the body is leaning back, every step tends to hit harder and slow you down more abruptly. A better cue is to stay over the feet, keep the steps short enough that you can still change them, and let the descent come through you instead of trying to pin it down one step at a time.

On a steeper downhill, the main question is no longer how much speed you can carry. It is whether each landing stays controlled. The right decision may be to stop attacking the descent and come down with shorter steps, more rhythm, and less last-second slowing than the runner next to you.

Technical descents

Here the strongest practical cue is where you look. In many situations you want the eyes a couple of footfalls, or roughly two to three metres, ahead rather than fixed right under your shoes. Do not be afraid to look that far ahead. You are not giving up control. You are giving the brain a little more time to organise the next step or two while peripheral vision still helps you track what is happening closer in. Looking only at your feet makes you late, which usually means more last-second slowing, more hesitation, and a higher nervous-system cost on top of the mechanical one.

On technical terrain, downhill running becomes a foot-placement task. The best line is not always the straight one. The fastest step is not always the longest one. The job is to stay composed enough that the next two or three landings are still deliberate.

Loose gravel, mud, scree, and snow

Loose surfaces change the demand again. On gravel, mud, scree, or old snow you may not get the same solid footing you expect on dry dirt or rock. That means less certainty, less direct force transfer, and a bigger penalty for any stride that is too long or too aggressive.

On these surfaces, staying organised beats looking smooth. You want enough control that one slip does not turn into three corrective steps. In practice that means a shorter stride, slightly quicker rhythm, softer knees, and a more available upper body for balance. The goal is not to hold your normal downhill speed. It is to keep enough options that the next step can still change if the ground moves under you.

Mud punishes force. If you push too hard into the ground, the foot slides and the whole body has to correct. Scree punishes overcommitment. If you throw the body downhill and the surface moves more than expected, the next two or three steps become recovery steps. Old snow or loose snow punishes certainty. The surface may hold on one step and collapse on the next. Across all of them, the practical answer is similar: shorter steps, less force into each landing, and less attachment to one exact line. On this kind of terrain, control beats boldness.

When poles help on the descent, and when they do not

Pole use downhill depends on the terrain. On smoother, runnable descents poles often add more clutter than help. If the trail is open and the job is mostly rhythm, stride control, and carrying speed well, poles can get in the way of arm swing and quick adjustments.

On steeper descents, poles can help when they keep you balanced and less defensive. Used well, they can take a little load out of the landing and give you another point of contact when the slope gets steeper or less stable underfoot. Used poorly, they just become extra things to place, extra things to think about, and another reason the upper body tightens.

On technical descents, poles are context-dependent. They can add stability and buy time on uneven or uncertain steps, but they can also occupy the hands when you need them free for balance, quick corrections, or a small catch on rock. On loose gravel, scree, or snow, they help only if they keep you centred without tangling your rhythm.

There are also a couple of practical cautions worth knowing. On snow, poles can help with balance, but a caught tip can jam, overload, or break while the body keeps moving downhill. On technical descents, many runners keep their hands easy to release from straps or glove attachments so a missed plant or small fall does not turn the pole into something the upper body has to fight.

A simple rule: use poles downhill when they clearly improve balance and control. Put them away when they slow decisions down or make you stiffer. Train the exact job you plan to ask of them in a race: when to plant, when to carry, and when to get them out of the way.

How do skyrunners move like that?

If you have watched skyrunners come down steep, technical terrain at a pace that looks almost unreasonable, it is easy to think the main difference is fearlessness. Most of the time it is something more concrete and more trainable than that.

What skyrunners tend to have is a much more trained relationship with unstable terrain. They read the surface earlier. They trust their feet to correct a small mistake without turning it into panic. They do not need every step to feel fully certain before they commit to moving. That is different from not caring about falling. It comes from seeing the same kind of terrain challenge often enough that the body and brain stop treating every small loss of certainty as an emergency and start learning how to handle it.

Courage is only part of it. Exposure, reactive balance, fast pattern recognition, and trust in your own corrections count just as much. The best descenders are not running blind. They are seeing more, deciding earlier, and hesitating less. Even when they do slip a little, the body is ready to catch it before it becomes a full mistake.

Confidence on technical descents improves best through progressive exposure, not through pep talks. Start with terrain that makes you pay attention without overwhelming you. Repeat it enough that the body begins to trust what it can correct. Then make the challenge a little harder. That is how downhill confidence becomes real: the nervous system gathers enough evidence that it can handle what is coming.

Rolling terrain versus one long descent

Course profile changes the job more than many runners realise. Rolling terrain keeps changing the task. You climb a little, descend a little, change rhythm, change loading pattern, and give some structures at least brief moments of relief. That means you can be a little more dynamic, because the job keeps changing before one system gets loaded the same way for too long.

One long uninterrupted descent is different. There is less reset. The same landing pattern keeps repeating, the same tissues keep taking the load, and the same weak link keeps getting tested. That is where you can feel good at the top, still move well for a while, and then suddenly find that one structure has become the limiter: quads, lateral knee, calves, feet, or technical ankle control.

Practically, that changes both training and pacing. If your race is mostly rolling, you need general downhill competence and the ability to keep switching well between climbing and descending. If the race is built around one long defining descent, train for that exact problem. And on race day, be more conservative in the first part of that downhill than your fresh legs want to be.

Late-race descents

Late in a long race, the legs are no longer the same legs you had in hour two. That is where a lot of people get caught. They try to run the downhill they could have run earlier. Meanwhile the quads are duller, the corrections are slower, and the same descent starts costing more than it did before.

At this stage, downhill running becomes less about gaining time and more about protecting control. Shorten the stride again, give each landing more room, and stop chasing the descent your fresh legs could handle. A little less aggression can keep the rest of the race from falling apart faster.

Gaze still matters. Keep the eyes a few steps ahead on technical ground and a bit farther out on faster open terrain. If fatigue pulls your vision down toward your feet, line choice gets later and the whole descent becomes more reactive.

Train the legs, train the skill

Many runners train uphill-specific work and leave descents to chance. That works until the races get long, steep, or technical enough to expose the gap.

Downhill improvement tends to come from three ingredients working together: eccentric tolerance, specific descent exposure, and the skill to keep making good foot placements when the trail gets less predictable.

Build eccentric tolerance on purpose

The repeated-bout effect is one of the strongest practical ideas in downhill science for runners. In simple terms, one dose of eccentric downhill work can make the next one less destructive for weeks. Downhill tolerance improves best through repeated exposure, not through one huge session.

In practice, start with progressive, pain-free eccentric strength for the quadriceps and lower leg: slow split squats, eccentric step-downs, slow heel drops, and controlled calf work. Then support that with manageable downhill running or downhill hiking volume. A strong build often includes 8-12 weeks where this work is present consistently instead of appearing once in a heroic session.

If descents always fall apart in the same place, train that place first. If the quads always go, build more quadriceps eccentric tolerance. If the calves are always the limiter, stop pretending that all downhill strength is a knee problem. The point is not just to get stronger in general. It is to make the first weak link less easy to expose.

Give the nervous system real descents

Strength helps, but it does not replace terrain exposure. You still have to learn how a slightly quicker cadence works on a runnable trail, what line choice does on loose rock, and how your body behaves when the slope changes under fatigue.

Downhill training therefore needs actual downhill running. Smooth downhill repeats, technical descents at low fatigue, back-to-back climb-and-descend blocks, and race-like terrain all belong here. Tougher quads matter, and so do cleaner decisions at speed.

There is also a practical sequencing point here. If you want to learn line choice, ankle control, and downhill rhythm, do some of that work while you are still fresh enough to actually learn. If you want to know what your form and decision-making do late in a race, add some descents after climbing or later in a long run. Both belong in training. They just answer different questions.

If a certain kind of downhill keeps slowing you down, train that kind of downhill. Smooth fast descents reward rhythm and landing control. Steep technical descents reward foot placement, gaze, and confidence under changing terrain. Loose gravel, scree, and snow reward adaptability and staying relaxed enough to keep options. The most direct preparation is still specific exposure, if you have access to it.

If you do not, you can still get part of the way there. Eccentric strength, balance work, quick-feet drills, and short technical repetitions on whatever uneven terrain you do have are all better than hoping race day will teach the lesson for free. But if the goal race is defined by one particular downhill problem, the best preparation is to meet that problem in training before it matters.

Do not hide the recovery cost from yourself

Hard descent sessions can flatten the rest of a week if you drop them into the plan carelessly. The legs may feel only mildly worked right after, then much worse the next day. The delayed cost means downhill training has to be placed with the same respect you would give a hard interval session or a big long run.

A better habit is to stop judging downhill training by the downhill pace alone and look at what it leaves behind over the next 48-96 hours. If the session improves your tolerance and your technique, good. If it wrecks two more training days every time, the dose is wrong.

The same goes for where the load shows up. If every hard descent beats up the quads, that tells you one thing. If the first limit is always the calves, feet, or ankle control on technical ground, that tells you another. Getting sore and calling it specific misses the point. Learn where your downhill system is weakest and train that part on purpose.

Keep one race-specific downhill stimulus in the build

For mountain races, a substantial downhill session or race-specific descent block often fits 14-21 days before race day. That is close enough to refresh the repeated-bout effect and far enough to let the damage settle. Put it too late and you arrive sore. Skip it entirely and race day may become the first time your legs meet the real cost of the descents.

This is even more important when the course has one long defining descent instead of rolling ups and downs. If the profile concentrates the downhill cost into one extended section, the preparation should reflect that. The strategy should too.

The most common mistakes

Training climbs but barely training descents. Uphill fitness does not automatically give you downhill tolerance or downhill skill. If you want to move well downhill late in a race, the descents need to be trained on purpose. Put real downhill work into the build instead of hoping race day will cover the gap.

Using one savage downhill session as race prep. Soreness is not the same thing as adaptation. Repeated, manageable exposure works better than one session that wrecks the next few days and teaches little besides how to get sore. Build tolerance through repeated doses the legs can actually absorb.

Leaving the last hard downhill session too close to race day. The final substantial downhill stimulus usually belongs roughly two to three weeks out, not right before the race. Close enough to refresh the legs, far enough to let the damage settle. Get the last big downhill session done early enough that you arrive recovered.

Treating early descents as free speed. Fresh legs make that easy to justify. The problem is that the bill usually arrives later, when the next climb costs more, the quads stop giving back, and technical ground starts needing more attention than it should. Start the first meaningful descent more conservatively than your fresh legs want to.

Running every downhill as if it asks for the same answer. A smooth runnable descent, a steep chute, late-race legs, and loose mud or scree do not reward the same kind of running. The more the terrain moves, steepens, or gets technical, the less this is about holding speed and the more it is about keeping options. Match the descent you are on instead of forcing one style everywhere.

FAQ

References

• Gottschall JS, Kram R. Ground reaction forces during downhill and uphill running. Journal of Biomechanics. 2005.
• Minetti AE et al. Energy cost of walking and running at extreme uphill and downhill slopes. Journal of Applied Physiology. 2002.
• Vernillo G et al. Biomechanics of graded running, Part I and II. Scandinavian Journal of Medicine & Science in Sports. 2020.
• Bjorklund G et al. Performance variability across uphill, downhill, and flat sections in trail running. Frontiers in Physiology. 2019.
• Bettega A et al. Time loss and speed determinants on ascents and descents in trail runners. International Journal of Sports Science & Coaching. 2026.
• Giandolini M et al. Foot strike pattern, fatigue, and downhill trail running mechanics. Journal of Biomechanics. 2016; Scandinavian Journal of Medicine & Science in Sports. 2017.
• Millet GY et al. Alterations of neuromuscular function after an extreme mountain ultramarathon. PLoS One. 2011.
• Heiderscheit BC et al. Effects of step rate manipulation on joint mechanics during running. Medicine & Science in Sports & Exercise. 2011.
• Lenhart RL et al. Increasing running step rate reduces patellofemoral joint forces. Medicine & Science in Sports & Exercise. 2014.
• Ho KY et al. Decline running increases patellofemoral joint stress compared with level and incline running. Physical Therapy in Sport. 2019.
• Bontemps B et al. Downhill running: what are the effects and how can we adapt? Sports Medicine. 2020.
• Hupperets MDW, Verhagen EALM, van Mechelen W. Effect of unsupervised home-based proprioceptive training on recurrences of ankle sprain. BMJ. 2009.