Cross-training for cyclists: running, skiing and swimming that transfer

Every winter the same question comes up: does the running, skiing or swimming I do when I cannot ride actually make me a better cyclist? The honest answer is partly. Fitness you build in another sport transfers to the bike through your heart, lungs and blood, and that part is real. It does not transfer through your legs, because the muscles, the joint angles and the nerve patterns that make a cyclist fast are specific to pedalling. This post sets out what the research says about that split, ranks the three common alternatives by how much of their work reaches the bike, and says when cross-training is a good trade and when it is simply a way of not training.

What "transfer" means and how much of it there is

The clearest summary is still Tanaka's 1994 review of cross-training in Sports Medicine. Its conclusion was that training in one endurance mode improves maximal oxygen uptake measured in another mode, but by less than training in that mode would have, and that the closer the two activities are in the muscles they use, the more of the gain carries across. Running and cycling, which both lean on the big leg extensors, transfer to each other reasonably well. Swimming, which is mostly arms and torso, transfers much less to either.

The classic demonstration is older. Pechar and colleagues trained people either on a bicycle or on a treadmill and then tested them both ways. Cycle training raised cycling VO₂max more than treadmill VO₂max; treadmill training raised treadmill VO₂max more than cycling. Each group got a general improvement and a larger specific one. That pattern, a general cardiovascular gain plus a specific muscular gain, is the whole story of cross-training in one experiment.

Foster's group ran the practical version with runners. One group added more running, another added swimming instead, and both were tested on the track afterwards. Both improved, which is the argument for cross-training; the run-trained group improved more, which is the argument against relying on it. In elite triathletes, who train all three sports at once, Millet and colleagues modelled how each discipline's training predicted performance in the others and found measurable transfer between cycling and running, with swimming largely staying in its own lane.

Why the legs do not come along

Part of what makes you a cyclist lives in the heart: a bigger stroke volume, more blood, more capillaries. That equipment does not care which sport built it. The other part lives in the muscle fibres that do the pedalling: mitochondrial density, the enzymes that burn fat and carbohydrate, and the coordination pattern that lets you push 250 W at 90 rpm without wasting any of it. Those adaptations are local. Running strengthens the calf and the hip in a way cycling does not need, and neglects the sustained knee-extension that cycling is made of. Cross-country skiing uses the quadriceps hard but in a different range of motion and with the upper body contributing a large share. Swimming barely touches the legs' endurance at all.

This is why a fit runner who takes up cycling feels their lungs are fine and their legs are not, and why a cyclist who tries a 10 km run is left with sore calves and a surprisingly ordinary time. The cardiovascular engine is shared. The drivetrain is not.

Ranking the three

• Cross-country skiing is the best of the three. It is whole-body, weight-bearing without impact, and it demands more oxygen than almost any other sport: the highest VO₂max values ever recorded belong to skiers, and Holmberg's review describes why, with the upper body adding a large share of the work at race intensity. What makes an elite skier fast reads like a list of what makes a cyclist fast, with the addition of arms. A long day of classic skiing is a long aerobic session for the legs and heart alike. The obvious limitation is that most cyclists do not live near snow.
• Running is the most available and the most dangerous. The cardiovascular return is excellent and it takes forty minutes, not three hours. But cyclists have not built the tendons and bones that running loads, and novice runners get hurt at a rate that surprises people. In Buist's randomised trial, roughly one in five novice runners picked up a running injury in thirteen weeks, and a gentler build-up did not reduce that number. Cyclists starting to run should assume they are novices, however fit they are.
• Swimming keeps the heart honest and does almost nothing for the legs. It is a good choice during an injury that rules out weight-bearing, and a reasonable choice for recovery. As a way of getting faster on the bike it is the weakest of the three, which Tanaka's review and the triathlon data both support.

The one benefit that has nothing to do with speed

Cyclists, like swimmers, tend to have lower bone density than runners, because pedalling loads the skeleton very little. Scofield and Hecht's review of bone health in endurance athletes sets this out, and it is the strongest argument for adding some impact work over a career even if it never made you a watt faster. Running is the simplest form of that, as is skiing. We cover the topic in more depth in bone density in cyclists. If you are choosing between running and swimming purely on health grounds, running wins.

Strength work is a different question

People often lump the gym in with cross-training. It is not the same thing. Heavy strength training does not build aerobic fitness at all, but Rønnestad and Mujika's review found that it improves cycling economy and performance in trained riders when it is programmed sensibly, largely through neural and muscle-fibre changes rather than through the heart. It is complementary, not a substitute, and we have a full guide in strength training for cyclists. The interference question, whether the gym blunts endurance gains, is covered in concurrent training interference.

When cross-training is the right call

• When you cannot ride. Injury, travel, ice on the roads, no trainer. An hour of running or skiing keeps the cardiovascular side from sliding and is far better than nothing. The rate at which fitness goes when you stop entirely is in detraining: how fast you lose fitness, and it is quicker than most people expect.
• When motivation is the limiting factor. A rider who would otherwise skip the session is better off running it. The physiology does not care that it was the second-best choice.
• When you want bone and connective-tissue health. Some impact work, kept small and built up slowly, is a good long-term investment.

When it is not

• In the twelve weeks before a target event. This is the period where specificity matters most. Every hour spent on a different sport is an hour of leg-specific adaptation you did not get. If the roads are unrideable, the trainer is the right substitute, not the pool.
• As the main source of volume. Two rides and three runs a week will make you a better runner. The cycling data in Pechar's and Foster's studies are clear that the specific mode wins, and there is no reason to think the reverse direction is different.
• As a way to sneak in more hard work. The heart does not know you were running. A hard run the day after a hard ride is two hard days, and the recovery cost is the same. Count it in your weekly load.

How we would split a winter week

These are our own working figures, not something with a study behind them. For a rider with six to eight hours a week who cannot ride outside from December to February, we would keep roughly two-thirds of the time on the bike, indoors, with the intervals there, and put the remaining third into skiing if it is available, otherwise into running built up from twenty-minute jogs over six weeks or more. Swimming would take the place of an easy day rather than a training one. The bike sessions are what the plan should be built around; the cross-training fills in around them.

If you are building a plan in Moveee, the plan wizard asks for the hours and days you actually have for riding, and the bike sessions it schedules are the ones worth protecting. Rides synced from Strava, Garmin or Wahoo are matched against those sessions on the calendar; runs and swims are not, because the engine only counts cycling load, which is the correct thing for it to do. The Indoor app covers the days when the roads are not an option and the plan still wants intervals.

The short version

Cross-training moves your heart and lungs forward and leaves your legs roughly where they were. Skiing is the best alternative, running the most practical and the most likely to hurt you, swimming the least useful for speed. Use it when the bike is not available, keep it out of the last block before a goal, and count its hard days as hard days.