The short answer: yes, sometimes. The longer answer involves knowing when cycling is a smart substitution and when it's a convenient excuse to skip the work your legs actually need.
Every marathon build hits this moment. Week nine, your calf is tight, the forecast says 34°C, and Sunday's 30km long run is staring at you from the training plan. The bike in the garage starts to look like a loophole. Same heart rate, same duration, none of the pounding. Surely the engine doesn't care how it gets trained?
Partly right. And the part that's wrong is precisely the part that decides whether you run 3:29 or walk the last 10km.
What the research actually says
In 2026, Menges and colleagues published a meta-analysis pulling together the controlled studies on substituting running with cycling. The headline finding: when athletes replaced a portion of their running volume with cycling, VO2max was preserved. No statistically significant difference against the run-only groups.
That's a genuinely useful result. It confirms what exercise physiologists have argued since the 1970s — the central cardiovascular system doesn't much care what movement pattern raises your heart rate. Cardiac output, plasma volume, mitochondrial signalling: all of it responds to sustained aerobic work, on foot or on wheels. I've covered the mechanism in detail in the fitness transfer guide.
But read the methods before you sell your running shoes. The studies in that meta-analysis replaced 20-50% of running volume. Not 100%. Nobody has run a controlled trial where marathoners did all their training on a bike and then raced 42km, for the obvious reason that an ethics board would laugh it out of the room.
There's a second detail worth pulling out of the methods. Most of these interventions ran for weeks, not seasons, and they measured VO2max — the size of the engine — rather than marathon performance itself. VO2max is the easiest fitness quality to preserve and the least specific one. A study can show your engine held steady while saying nothing about whether your legs would survive the last hour of a marathon. That's not a criticism of the research; it's a limit on what the research can tell you.
So the straight reading is this: partial substitution preserves the aerobic engine. Full substitution is untested, and everything we know about tissue adaptation says it would go badly.
RunnersConnect — one of the more rigorous coaching outfits in the running world — lands on a similar number from the practical side: cycling can absorb about 15-20% of running volume without a measurable loss in running performance. Beyond that, they flag a cost the VO2max studies don't capture: a 3-5% running economy penalty when cross-training displaces running-specific work.
Three to five per cent sounds small. Over a marathon, it's somewhere between eight and fifteen minutes. It's the difference between a Boston qualifier and a near miss.
Why economy is the sticking point
VO2max is the size of your engine. Running economy is how much fuel that engine burns to hold a given pace. And economy, unlike VO2max, is stubbornly sport-specific.
Running economy lives in three places the bike never touches.
The stretch-shortening cycle. Every stride, your Achilles tendon and calf load up like a spring during landing and return that energy at push-off. A well-trained runner gets a meaningful share of each stride for free — elastic recoil, not muscular work. That spring stiffens and sharpens only through repeated impact. Cycling is concentric-only: the pedal never asks your tendons to absorb and return anything.
Eccentric loading. Running is controlled falling. Every footstrike, your quads and calves lengthen under load to brake 2-3 times your body weight. That eccentric conditioning is what lets you still run properly at kilometre 35 when the accumulated damage would fold an untrained leg. There is no eccentric phase in pedalling. None.
Neuromuscular coordination. Cadence, ground contact time, hip stability at 180 single-leg landings per minute — a motor skill, rehearsed into efficiency. Time on the bike rehearses a different skill.
This is why the transfer between the two sports is asymmetric and why a 300-watt cyclist gets humbled in a parkrun. The engine transfers. The drivetrain doesn't.
When cycling can replace a run
None of this makes the bike useless in a marathon build. Used in the right slots, it's one of the smartest tools available. Four situations where the substitution genuinely works:
Injury management. A grumbling Achilles, a hot spot on the shin, plantar fascia that argues for the first ten minutes of every run. These are the moments where one more run turns a niggle into a six-week layoff. A bike session holds your aerobic load at full volume while the tissue calms down. This is the difference between missing three days and missing a month, and it's the single best argument for every marathoner keeping a bike in the garage. The same logic applies in reverse for cyclists adding run volume — the injury prevention guide covers that direction.
Excessive fatigue. Marathon training accumulates structural damage faster than aerobic fatigue. If your legs feel like wood but your motivation and heart rate are normal, the limiting factor is tissue, not engine. Swapping an easy run for an easy ride delivers the aerobic stimulus while the pounding stops. Recovery continues, fitness doesn't leak.
Heat. Running in 34°C at marathon-training volumes is a losing proposition — pace collapses, cardiac drift wrecks the intended stimulus, and the recovery cost balloons. An indoor trainer with a fan gives you a controlled environment and an honest session on a day when running would deliver neither.
Time-crunched days. An hour's trainer session starts thirty seconds after you decide to do it. No drive to a route, no traffic lights, no faffing with kit. On the days when the choice is a compromised 40-minute run or a proper 60-minute ride, the ride often wins — the time-crunched running-cycling combination works in both directions.
Notice the pattern: in every one of these, cycling replaces an easy or moderate run. Aerobic volume in, impact cost out.
When it can't
Now the other list — the sessions where the bike is a false economy.
The long run in the final 6-8 weeks. This is the one runners most want to swap and the one that least tolerates it. In the last two months before a marathon, the long run's primary job quietly changes. Early in the build it's mostly an aerobic session. By the sharpening phase, it's a structural one: teaching your muscles, tendons, bones, and connective tissue to tolerate two, three, four hours of repeated impact while progressively fatigued. There's a phrase for what happens to runners who arrive at race day with a big engine and unprepared tissue — the last 10K of a marathon is full of them. Aerobically fine, structurally finished, watching the pace slip away one kilometre at a time.
You need the pounding. That sentence is unfashionable and true. Tissue tolerance to impact is built by progressive exposure to impact, and there is no cycling workaround.
Running-specific intensity. Tempo runs, threshold sessions, marathon-pace blocks — these train economy at race pace, rehearsing the exact neuromuscular pattern you'll use on the day. A threshold session on the bike trains your engine and your quads' pedalling endurance. Useful for cycling. Nearly worthless for locking in what 4:45/km feels like at kilometre 30.
Race-specific preparation generally. The closer you get to the marathon, the more your training needs to look like the marathon. This is the oldest principle in the sport, and it survives every new study because it keeps being right.
The framework: which runs are actually replaceable
Sort your week into three tiers and the decision makes itself.
Tier 1 — never substitute. The long run inside 6-8 weeks of race day. Marathon-pace and threshold sessions. These are the runs your race is built on.
Tier 2 — substitute when needed. Easy runs, recovery runs, general aerobic volume. This is where the 15-20% lives. A midweek easy 50 minutes becomes a 60-75 minute Zone 2 ride and the training week barely notices.
Tier 3 — substitute freely. Anything you'd otherwise skip. The junk-fatigue day, the travel day, the day the physio said no running. A bike session here isn't replacing a run — it's replacing a zero.
Run the arithmetic on a typical 70km week: roughly 45-50km of it is easy running. Moving 10-14km of that to the bike costs you nothing measurable and buys you a week of fresher tendons. Moving the Sunday 32km to the bike in week fourteen costs you the exact adaptation the race demands.
What the swap looks like in practice
Distance doesn't convert cleanly between the sports, so when you do substitute, swap by time and effort rather than kilometres. A 50-minute easy run becomes a 60-75 minute Zone 2 ride — conversational effort, higher cadence, no surges. The full exchange-rate maths is its own topic, covered in the conversion guide, but the short version is that cycling needs roughly one and a half times the duration to match a run's aerobic stimulus, because the bike carries your body weight and running doesn't.
Here's a marathon-build week, week eight of sixteen, with the substitution done properly:
Monday, rest. Tuesday, running intervals — untouchable, this is race-specific work. Wednesday, 60-75 minutes easy on the bike, replacing the usual 50-minute recovery run because Tuesday's session left the calves tender. Thursday, tempo run — untouchable. Friday, rest or 45 minutes very easy spinning. Saturday, easy run, kept on foot because the week needs its impact exposure. Sunday, the long run, on foot, always on foot from here to the taper.
One substitution, maybe two in a heavy week. The running skeleton of the week — intervals, tempo, long run — stays intact, and the bike quietly absorbs the volume that was only ever there for aerobic maintenance. That's the entire trick. The riders and runners who get this wrong aren't the ones who swap a Wednesday recovery run; they're the ones who let the bike creep into the sessions that define the race.
One more practical note: keep the substituted rides genuinely easy. The temptation on a bike is to chase speed, chase segments, half-race the group that passes you. A recovery-run replacement ridden at tempo isn't a substitution — it's an extra hard day smuggled into your week, and it will show up as flat legs on Thursday.
Where that leaves you
Cycling is excellent insurance and excellent maintenance. It keeps the engine running when your structure needs a break, absorbs aerobic volume with near-zero injury risk, and turns forced rest days into training days. Runners who use it that way get through marathon builds with fewer interruptions, and consistency beats any single session ever invented.
But if the goal is a fast marathon, you need to run. Economy is built on foot. Tissue tolerance is built on foot. The final six weeks are built on foot. The bike protects the work; it doesn't do the work.
And if you don't have a race on the calendar? Then the calculus flips. For staying fit, healthy, and durable across a year, cycling can absorb a significant chunk of your aerobic volume — arguably it should, since it delivers the same engine work at a fraction of the orthopaedic cost. Plenty of runners in their forties and fifties have quietly discovered that a 60/40 run-bike split keeps them training every week of the year instead of cycling through injury layoffs. The full playbook for combining the two is in the running and cross-training guide, and the weekly scheduling guide shows how to slot the sessions together without wrecking either sport.
Know which question you're asking. "Can cycling keep me fit?" Absolutely. "Can cycling make me a faster marathoner without running?" No — and anyone selling you otherwise hasn't stood at kilometre 35 and watched the difference walk past.