Which way should you ride? Prevailing winds on great cycling routes
Find the direction the wind favours on 15 famous cycling routes and three windy places, by season and by month, from ten years of data.
The short answer
Follow the prevailing wind direction where cycling routes run one way: north on LEJOG, south on the Pacific Coast, east across America and down the Danube. Ten years of hourly data on 15 routes show the gain is real but modest, and on loops such as Iceland's Ring Road it disappears. Your own dates still need a forecast.
Key takeaways
- Ten of our 15 routes had a better direction; on the other five, the two ways were within 8 points.
- The mid-latitude westerlies explain most verdicts, which is why east and north so often win.
- On a loop, a steady wind helps on one side as much as it hinders on the other.
- The Pacific Coast had the widest gap, with 17% headwind hours southbound against 52% riding north.
- Prevailing isn't guaranteed, so check a route-timed forecast for your own dates.
Definitions
- Prevailing wind
- The direction the wind blows from most often at a place, over many years.
- Along-route wind
- The part of the wind blowing straight along your direction of travel; against you it's a headwind, behind you a tailwind.
- Headwind hour
- A riding hour in which more than 4 km/h (2.5 mph) of the wind blew against your direction of travel.
- ERA5 reanalysis
- A record of past weather that blends observations from around the world with a weather model to fill in every hour and place.
Which direction should you ride, going by the prevailing wind?
Ride with the prevailing wind direction where a cycling route runs one way, and stop worrying about it on a loop. We tested that against ten years of hourly wind on 15 famous routes on four continents, in both directions, plus three famously windy places.
On ten routes, one direction clearly had the better wind. On the other five, the two ways were within 8 percentage points of headwind hours, so we call them "either way". The push is usually modest: a tiebreaker worth hours over a long tour, not days. It still counts, because a headwind costs more than a tailwind gives back.
This page sums up the study. Each route has its own guide with months, sections and years.
What are prevailing winds, and why do they have a direction?
They're the winds a place gets most often, set by the planet's big circulation belts and bent by coasts and mountains.
- The westerlies. NOAA puts westerly winds between about 35° and 60° north and south. Most of Europe, the northern US and southern Australia and New Zealand sit in them, which is why so many of our verdicts point east.
- The trade winds. Between about 30° north and south, NOAA explains, they blow from the north-east in the northern hemisphere and the south-east in the southern. None of our routes lies in them.
- Sea breezes. By day, land warms faster than the sea, so cooler sea air flows inland, as NOAA describes. On coasts that can matter more than the big belts: Mallorca's summer wind swings onshore by mid-afternoon.
- Local winds. Mountains and valleys make their own. The mistral blows down the Rhône valley past Mont Ventoux; in Kansas, southerlies dominate summer.
Want this for your own route? Windline shows the wind for every stretch, timed to your pace. Free. Make a free ride page →
How did we work out the better direction?
We scored every riding hour of ten years against the road, in both directions.
- The data. Hourly wind 10 m (33 ft) above the ground from ERA5, the reanalysis that ECMWF produces for the Copernicus Climate Change Service, for 2015 to 2024.
- The route. A line along each route's usual course, taking the wind from every grid cell it crosses, about every 25 km (16 miles).
- Riding hours. 08:00 to 17:00 local time, in the months most people ride.
- Headwind or tailwind. We split each hour's wind into the part along the road and the part across it. More than 4 km/h (2.5 mph) against you is a headwind hour; the same behind you is a tailwind hour; the rest is side-on or calm.
- The verdict. The direction with at least 8 points fewer headwind hours wins. Closer than that, it's either way.
The full method is in the sources section. ERA5 smooths headlands, valleys and summits, so treat the figures as habits, not roadside readings.
Which way should you ride each route?
Here's the whole study in one table, with the better direction's headwind hours against the other way's.
| Route | Season | Wind most often from (riding hours) | Better direction | Headwind hours, better way / other way |
|---|---|---|---|---|
| Land's End to John o' Groats, UK | April to September | West 19.2%, south-west 18.9% | North (LEJOG) | 29.4% / 44.4% |
| Hebridean Way, Scotland | May to September | South-west 20.3%, south 19.7% | North, Vatersay to the Butt of Lewis | 31.0% / 55.1% |
| North Coast 500, Scotland | May to September | South-west 22.0%, west 15.7% | Either way | Clockwise 38.4%, anticlockwise 38.0% |
| Pacific Coast, Canada and USA | May to October | North-west 25.9%, west 22.3% | South | 16.6% / 52.0% |
| TransAmerica Trail, USA | May to September | West 16.1%, south-west 15.2% | East | 22.2% / 35.1% |
| Great Allegheny Passage and C&O Canal, USA | April to October | West 18.0%, north-west 16.4% | East, Pittsburgh to Washington, DC | 23.8% / 38.3% |
| Great Divide, Canada and USA | June to September | West 26.1%, south-west 21.9% | Either way | Southbound 27.1%, northbound 23.7% |
| Danube Cycle Path, to Budapest | April to October | North-west 15.4%, west 15.3% | Downstream | 27.5% / 42.4% |
| Rhine Cycle Route | May to September | South-west 16.4%, west 13.1% | Either way on wind; downstream for the gradient | Downstream 31.7%, upstream 24.7% |
| Loire à Vélo, France | April to October | South-west 18.8%, north-east 16.5% | Upstream, away from the Atlantic | 31.3% / 41.3% |
| Camino Francés, Spain | April to October | North-east 13.7%, north 13.0% | Either way | Westbound 29.8%, eastbound 31.1% |
| Iceland's Ring Road | June to August | North 14.6%, south-west 12.9% | Either way | Clockwise 36.0%, anticlockwise 34.6% |
| Otago Central Rail Trail, New Zealand | October to April | North-west 28.5%, west 14.0% | East, Clyde to Middlemarch | 16.1% / 38.5% |
| Great Ocean Road, Australia | October to April | South-west 18.2%, west 17.7% | East, Allansford to Torquay | 34.1% / 45.6% |
| Shimanami Kaido, Japan | March to May, October to November | North-east 24.2%, south-west 15.6% | South, Onomichi to Imabari | 29.6% / 39.4% |
Three places are about when to ride, not which way:
| Place | Wind most often from | What the data says |
|---|---|---|
| The Netherlands, six places | South-west 25.2%, west 16.4% | Ride east or north-east where you can; a loop spends 48.4% of riding time in a headwind |
| Mallorca, Spain | Days with east wind 31.3%, west 28.9% | May balances wind, warmth and rain best; choose the loop by the day's wind |
| Mont Ventoux, France | North 25.5%, north-west 14.3% over the summit | June to September has the fewest gusty days; start early |
Which way should you ride in Britain?
North, on both of Britain's classic one-way routes. The south-westerlies push you up the country.
LEJOG or JOGLE: northbound riders averaged a 2.6 km/h (1.6 mph) tailwind from April to September, in all ten seasons. August helped most; April was close to a draw. Our model rider saved about four hours of pedalling over two weeks. Everything else about the ride is in our LEJOG route guide.
The Hebridean Way gave a bigger push: a 6.3 km/h (3.9 mph) average tailwind riding north from Vatersay. Southbound, 34.1% of riding hours had a strong headwind, over 15 km/h (9 mph). It's also exposed: side-on wind averaged 13.8 km/h (8.6 mph).
Which way should you ride across North America?
South down the Pacific Coast and east across the continent; on the Great Divide, it's either way.
- Pacific Coast. The summer north-westerlies gave southbound riders a 5.1 km/h (3.2 mph) tailwind from May to October, every year. July was best. In our 35-day trip test, south beat north from 93% of start dates.
- Across America. Eastbound had a 1.3 km/h (0.8 mph) tailwind on the TransAmerica Trail, every season. Kansas is the exception: its southerly winds favour riding west.
- Pittsburgh to DC. A 1.9 km/h (1.2 mph) average tailwind eastbound, in all ten seasons. Under trees along the rivers, heat and mud may matter more.
- The Great Divide. Westerlies blow across a route running north to south, so they're mostly side-on: 7.6 km/h (4.7 mph) on average. Northbound leans slightly ahead; snow and season decide more.
Which way should you ride Europe's rivers and the Camino?
Downstream on the Danube, upstream on the Loire, and on the Rhine and the Camino the wind barely cares.
- Danube. Downstream to Budapest averaged a 2.1 km/h (1.3 mph) tailwind, in eight of ten seasons. Riding upstream through Austria, the wind cost our rider over six times more than the gentle climb. Below Belgrade the advantage flips, and over the whole river to the Black Sea it's either way: 31.4% against 37.6%. See our Danube Cycle Path guide for trains home.
- Rhine. Downstream riders averaged a small headwind, 0.9 km/h (0.5 mph), in all ten seasons, but the gap in headwind hours was only 7 points, and 26.2% of riding hours were calm. Riding up from the North Sea adds a lot of climbing, so downstream wins.
- Loire. Riding east, away from the Atlantic, averaged a 1.6 km/h (1.0 mph) tailwind, in nine seasons of ten. Many riders head for the sea anyway; at these speeds, trains and dates can decide.
- Camino Francés. Westbound pilgrims met a headwind in 29.8% of hours, eastbound riders 31.1%: a draw.
Does direction matter on a loop or an island?
Much less than on a one-way route. A loop turns through every compass point, so a steady wind is behind you on one side and against you on the other.
Iceland's Ring Road shows it best. Clockwise and anticlockwise both averaged no push at all, and in our 14-day trip test clockwise won from 50.1% of start dates: a coin toss. Only the west, Reykjavík to Blönduós, leans one way, towards anticlockwise.
The North Coast 500 was the same: 38.4% headwind hours clockwise against 38.0% anticlockwise. Clockwise leans very slightly ahead, but decide on the morning from the forecast.
In the Netherlands, a loop met a headwind on 39.5% of the distance but 48.4% of the riding time, because you go slower into it. On point-to-point trips, finish in the east: in summer, riding east had a tailwind at all six places we tested.
When does the wind matter more than the direction?
On islands, mountains and coasts, where the question is when to ride rather than which way.
- Mallorca. May was the best balance: Palma's riding-hour wind averaged 12.4 km/h (7.7 mph), highs near 22°C (71°F) and the fewest wet days of the cycling months. The north coast around Pollença was windiest in every month, so on north-wind days the south coast is the calmer choice. Our Mallorca route guide has the loops.
- Mont Ventoux. The grid cell over the summit averaged 31.4 days a year with gusts of 70 km/h (43 mph) or more, most of them with mistral-sector winds. June to September had the fewest. The Vaucluse department advises against the climb in wind above 80 km/h (50 mph).
Which way should you ride in Australia, New Zealand and Japan?
East on both southern-hemisphere routes, and south on the Shimanami Kaido. The westerlies explain the first two.
- Great Ocean Road. From October to April, eastbound riders, Allansford to Torquay, averaged a 3.0 km/h (1.9 mph) tailwind and met a headwind in 34.1% of riding hours, against 45.6% riding west. East won in nine seasons of ten. The edge was smallest from January to March.
- Otago Central Rail Trail. Clyde to Middlemarch averaged a 2.6 km/h (1.6 mph) tailwind, in all ten seasons, with a headwind in 16.1% of riding hours against 38.5% riding west.
- Shimanami Kaido. Southbound, from Onomichi to Imabari, won clearly in October and November; spring was a draw.
Does the better direction change with the month?
On most routes the verdict holds through the riding season, but it's often a draw at the edges.
| Route | Better way, clearly | Close to a draw | Other way, clearly |
|---|---|---|---|
| LEJOG (north) | June to September | April, May | None |
| Hebridean Way (north) | May to September | None | None |
| Pacific Coast (south) | May to October | None | None |
| TransAmerica (east) | May to August | September | None |
| Pittsburgh to DC (east) | April, June to August, October | May, September | None |
| Great Divide | June, September northbound | July, August | None |
| Danube, whole river (downstream) | April, June, July | May, August to October | None |
| Rhine | None | August, September | May to July upstream |
| Loire (upstream) | June to September | April, May, October | None |
| Camino Francés | July westbound | April to June, August to October | None |
| Otago (east) | October to April | None | None |
| Great Ocean Road (east) | April, October to December | January to March | None |
| Shimanami Kaido (south) | October, November | March to May | None |
Iceland's Ring Road and the North Coast 500 were a draw in every month of their seasons.
Can you count on the prevailing wind for your trip?
For the averages, yes; for your fortnight, no. Even on the Pacific Coast, the clearest one-way verdict, southbound riders met a headwind in 16.6% of riding hours.
Use the study to pick a direction and month, then let the forecast take over.
- Pick the direction from the tables above, and the month where the verdict is clearest.
- Leave spare days, so one stormy day doesn't break the plan. See planning a multi-day cycling tour.
- Check a route-timed forecast, which gives each stretch's wind at the hour you'll reach it, a week out and the evening before. Reading a weather forecast for a bike ride explains why a town forecast isn't enough.
- Know your limits, and put official warnings first. Our guide to cycling in wind covers the rest.
Looking for your next route? See our lists of the most popular tours in Europe, North America, Australia and New Zealand and Africa.
Frequently asked questions
Why do so many cycling routes favour riding east?
Because most famous touring routes lie in the mid-latitudes, where the westerlies blow. NOAA places these westerly winds between about 35° and 60° north and south. A west wind pushes riders heading east, so the TransAmerica Trail, the Great Allegheny Passage and the Danube to Budapest all favoured eastbound riders. Local winds can still overrule the pattern, as Kansas's southerlies and the Loire's north-easterlies show.
Does the wind matter more than the hills when you choose a direction?
On flat river routes, usually yes. Riding the Danube upstream through Austria, the wind cost our model rider over six times more than the gentle climb. Mountains are different. The Rhine's summer wind leaned slightly upstream, but riding up from the North Sea adds about 1,200 m (3,940 ft) of climbing, so downstream still wins. Our [Danube direction guide](/blog/danube-cycle-path-direction/) shows the sums.
How much time does riding the better direction save?
Hours, not days, on most long routes. Our model of a loaded touring rider saved about four hours on Land's End to John o' Groats, about nine hours across America and roughly 20 hours on the Pacific Coast. The model rides at a steady 120 W and 24.1 km/h (15.0 mph) in still air, so a faster or lighter rider loses a little less time.
Which route in the study had the strongest wind advantage?
The Pacific Coast route, by headwind hours. From May to October, southbound riders met a headwind in 16.6% of riding hours, against 52.0% riding north: a 35-point gap. The Hebridean Way in Scotland had the biggest average push, a 6.3 km/h (3.9 mph) tailwind for northbound riders, on a far shorter route.
Is the prevailing wind on a route the same all year?
Not always. On the Pacific Coast, the southbound tailwind peaks in July and reverses in December and January. On the Great Ocean Road, eastbound riders' edge was smallest from January to March and largest from May to September, outside the usual season. Some patterns strengthen out of season: the Danube's downstream push and LEJOG's northbound one were both strongest in winter.
Why is the wind data measured 10 m above the ground?
That's the standard height for wind measurements, and the height ERA5 reports. Wind at saddle height is weaker, because the ground, hedges and buildings slow it. In our rider model we took the wind at rider height as 0.65 times the 10 m wind. The direction comparisons don't depend on that factor, because both directions face the same wind.
Does riding with the prevailing wind make a tour safer?
Not by itself. Side-on wind was often bigger than the push along the road: on LEJOG it averaged four times the size. Gusts from the side move riders around in traffic, whichever way they're heading. Direction choice saves time; safety comes from checking gusts and official warnings each day. Our guide to [how windy is too windy to cycle](/blog/how-windy-is-too-windy-to-cycle/) gives practical limits.
Can you use prevailing winds to plan a day ride or club loop?
Only loosely. On a loop that comes back to its start, a steady wind costs and repays in turn, so the long-term average says little about one morning. What matters is the wind on each stretch at the hour you reach it. For a single day, a forecast for that day beats any ten-year average.
Why use ten years of wind data rather than one?
Because single years vary. The Danube's downstream advantage held in eight of ten seasons from 2015 to 2024; in 2018 the wind leaned slightly the other way. Ten years shows whether a pattern is a habit or a fluke. On LEJOG, the Pacific Coast and the TransAmerica Trail, the better direction won in all ten seasons.
Get the wind for your ride.
Upload a GPX, add your start time and pace, and see the headwind, gusts, rain and heat for every stretch of the route, timed to when you’ll be there. Free for riders.
Make a free ride page →Sources and method
- Copernicus Climate Change Service (ECMWF) (2026). ERA5 hourly time-series data on single levels from 1940 to present
- Quarterly Journal of the Royal Meteorological Society (2020). The ERA5 global reanalysis (Hersbach et al.)
- ECMWF (2026). ERA5: data documentation
- NOAA National Ocean Service (2026). What are the trade winds?
- NOAA (National Oceanic and Atmospheric Administration) JetStream (2026). Global atmospheric circulations
- NOAA (National Oceanic and Atmospheric Administration) JetStream (2026). The Sea Breeze
Original analysis by Windline's data desk, October 2026, the same study behind each linked direction guide. Wind data: hourly 10 m wind (the u and v components, turned into speed and direction) and 10 m gusts from the Copernicus Climate Data Store, dataset ERA5 hourly time-series data on single levels (reanalysis-era5-single-levels-timeseries, doi:10.24381/1cf1ad76; ECMWF's ERA5 reanalysis for the Copernicus Climate Change Service), for 1 January 2015 to 31 December 2024 (ten full years). ERA5 is described by Hersbach et al. (2020). Routes: 15 lines through towns and road points following each route's usual course (LEJOG, the Pacific Coast route from Vancouver to Imperial Beach, the TransAmerica Trail, the Danube Cycle Path to the Black Sea, Iceland's Route 1, the North Coast 500, the Hebridean Way, the Rhine Cycle Route, the Loire à Vélo, the Camino Francés, the Great Divide Mountain Bike Route from Banff, the Great Allegheny Passage and C&O Canal, the Otago Central Rail Trail, the Great Ocean Road from Torquay to Allansford, and the Shimanami Kaido); none is an official route. Each line was cut into pieces of at most 25 km, and each piece takes its wind from the nearest ERA5 0.25° grid cell: every 0.25° grid cell along the route, about every 25 km, so results are weighted by distance. ECMWF gives the model's native resolution as about 31 km, so these are area averages, not roadside readings. Three site studies use single grid cells instead of a line: six places in the Netherlands, six cells over Mallorca, and four cells around Mont Ventoux and the Rhône valley. Riding hours are the values stamped 08:00 to 17:00 local time inclusive (10 a day), in each place's own time zone with daylight saving applied. Each route's season is the one named in the table, the months most riders ride it; monthly figures use each calendar month. For each hour and piece, the along-route wind is the wind speed times the cosine of the angle between where the wind comes from and the direction of travel; riding the other way flips the sign. A headwind hour has more than 4 km/h (2.5 mph) against you, a tailwind hour more than 4 km/h behind you; the rest are side-on or calm. "Wind most often from" counts where the wind came from in eight compass sectors over the season's riding hours, distance-weighted; hours under 5 km/h (3 mph) count as calm. Our verdict rule: a direction is "better" when its share of headwind hours is at least 8 percentage points lower than the other way's; closer than that, we call it "either way". The same rule sorts the months in the month table. The Danube verdict uses the Donaueschingen-to-Budapest section, where most people ride; the whole river is also given. Year-to-year figures are each season's average along-route wind, year by year. Trip tests ride each whole line in equal daily stages (14 days for LEJOG and Iceland, 35 for the Pacific Coast, 75 for the TransAmerica Trail, 30 for the Danube), from every start date in the season, and compare the two directions from the same start date. Modelled riding time uses a steady-power touring rider: 100 kg (220 lb) of rider, bike and luggage, 120 W, rolling resistance 0.006, drag area 0.45 m², air density 1.2 kg/m³, which gives 24.1 km/h (15.0 mph) in still air; wind at rider height is taken as 0.65 times the 10 m wind, and crosswind drag is ignored. The hours saved come from the linked guides: about 4 hours on LEJOG, 9 on the TransAmerica Trail, 20 on the Pacific Coast and two and a half on the Danube to Budapest. The Netherlands, Mallorca and Mont Ventoux figures are from the site studies and are explained in their guides. The Rhine's extra climbing upstream comes from SwitzerlandMobility, as cited in our Rhine Cycle Route guide. The westerly and trade-wind latitudes are NOAA's. Every other figure is in results.json in the study's data folder. Imperial conversions use 1 km/h = 0.621 mph, 1 mile = 1.609 km and 1 m = 3.28 ft. Contains modified Copernicus Climate Change Service information (ERA5, 2015–2024); neither the European Commission nor ECMWF is responsible for any use of this information.
About the author
Wes Thomson, Founder, Windline. Wes Thomson builds Windline, live wind and weather forecasts timed to when riders reach each stretch of road. He built it for his own 2,093 km tour from Pont Drift to Cape Town, and writes these guides with the same focus: what the wind will do to your ride, and how to plan around it.