What makes a good route
Before a route is ranked, we check its shape. Rides and runs face the same checks, and a route that fails any of them is left out, however green or quiet it is.
A good route brings you back to where you started without making you cover the same street twice. It heads out from the start rather than circling around it, so the start sits at its edge, not buried in the middle. It ranges widely, reaching out in many directions rather than running up one long corridor and back. And it flows: over any kilometre it keeps moving somewhere, rather than zigzagging through side streets to make up the distance.
How routes are ranked
Each route is measured along its own length: how much of it runs through nature, how quiet and how far from traffic it is, its longest unbroken stretch of nature, its views and how close it starts to a station. Rides also score their climbing. Rides and runs are weighted separately, by hand, and the weights are published here.
| Through nature | 30% |
|---|---|
| Views | 5% |
| Away from traffic | 30% |
| Longest nature stretch | 15% |
| Starts near a station | 20% |
| Through nature | 30% |
|---|---|
| Views | 5% |
| Away from traffic | 30% |
| Longest nature stretch | 15% |
| Climbing | 20% |
- Through nature
- How much of the route runs through parks, forests and along water.
- Views
- How much of it has open views.
- Away from traffic
- How much of it is on paths and streets with little or no car traffic.
- Longest nature stretch
- Its longest unbroken stretch of nature, as a share of the whole route. A gap shorter than 100 m does not break it.
- Starts near a station
- Full marks for starting right at a well connected station, falling to nothing 800 m away.
- Climbing
- Rides only: more climbing per km scores higher, up to 15 m per km.
How the best compare
To show how much better the routes on these pages are, we set them against a typical route: the one in the middle of the first route we build from every station and start, with half scoring higher and half lower. Only the starts that do well there get longer routes, so the middle of those first routes is the fair picture of the city. That is 80,566 runs and 2,891 rides in Perth.
Each row is one part of the score, out of 100%. Typical is the middle score of those first routes. On these pages is the middle score of the routes we chose: the best run in each district, and the best ride from each station at each distance.
| Typical | On these pages | |
|---|---|---|
| Through nature | 37% | 92% |
| Views | 0% | 6% |
| Away from traffic | 90% | 99% |
| Longest nature stretch | 17% | 80% |
| Starts near a station | 60% | 60% |
| Overall score | 52% | 83% |
| Typical | On these pages | |
|---|---|---|
| Through nature | 30% | 67% |
| Views | 0% | 18% |
| Away from traffic | 92% | 83% |
| Longest nature stretch | 10% | 29% |
| Climbing | 19% | 20% |
| Overall score | 42% | 47% |
| Without climbing, as cities are compared | 47% | 54% |
Where the routes start
Runs start at rail, metro and tram stations, spread out so they cover each district, with up to 10 per district. Each station gets a short test run, and the stations with the best test runs become starts.
Rides start at one station in each 3 km square of the city: a rail or metro station where there is one, else a tram stop, else any other station or bus stop, the one nearest the middle of the square. Each of those 253 stations gets a 20 km test ride, and the 13 with the best test rides, at least 6 km apart, become ride starts. From each we build rides of 20, 60 and 100 km.
Districts are the areas Perth is divided into, as mapped in OpenStreetMap.
| District | Stations tried | Starts kept |
|---|---|---|
| City of Canning | 621 | 10 |
| Town of Victoria Park | 235 | 10 |
| City of South Perth | 259 | 9 |
| City of Belmont | 279 | 9 |
| City of Gosnells | 640 | 10 |
| City of Melville | 462 | 10 |
| City of Perth | 215 | 8 |
| Kings Park | 6 | 1 |
| City of Vincent | 167 | 10 |
| City of Bayswater | 395 | 10 |
| Town of Bassendean | 101 | 8 |
| City of Subiaco | 75 | 9 |
| City of Nedlands | 154 | 10 |
| Town of Claremont | 60 | 9 |
| Town of East Fremantle | 31 | 6 |
| Town of Cambridge | 162 | 9 |
| Town of Mosman Park | 42 | 8 |
| Town of Cottesloe | 35 | 8 |
| City of Stirling | 1084 | 10 |
| City of Fremantle | 223 | 10 |
| City of Kalamunda | 414 | 10 |
| City of Armadale | 446 | 10 |
| City of Cockburn | 609 | 10 |
| City of Kwinana | 250 | 7 |
| City of Swan | 16 | 6 |
| Shire of Mundaring | 357 | 10 |
| City of Joondalup | 739 | 10 |
| Shire of Serpentine-Jarrahdale | 168 | 5 |
| City of Rockingham | 600 | 10 |
| City of Wanneroo | 947 | 10 |
| City of Mandurah | 468 | 10 |
After dark runs
An after dark run is lit along at least 60% of its whole length, going by the street lighting recorded in OpenStreetMap. Stretches with no record count as unlit, and a route only states how much of it is lit when lighting is recorded on at least 75% of it.
In Perth, lighting is recorded on 5% of the length of the runs we ranked. We offer after dark runs only above 50%, so there are none this month.
How long a route takes
Times assume a steady pace: 6 minutes per km running and 3 minutes per km riding. Your own pace will move them up or down.
Checked every month
Every route is built again each month on fresh map data, so a district's best run can change. This is the first month for Perth, so there is nothing to compare yet.
This edition
Built 24 October 2026 by engine 0a7bbbaba,
map data version 11, foot line catalog 16, cycle line catalog 16: 11,389 requests, 93,862 routes, of which 7,761 ranked, from 10,260 stations screened.