Run a 5 km loop around Midtown Manhattan and you cross about 34 traffic lights, one every 150 metres. Do the same in Ghent and you cross one or two. We had Runplanner's loop generator draw 1,000 running loops in 25 cities, 40 per city, and counted every traffic light on the way, at junctions and at pedestrian crossings: 8,483 lights over 5,036 km.
The short version
- New York is in a league of its own: 33.5 lights on a 5 km loop, nearly 50% more than second-placed Paris (22.9).
- Barcelona (21.5) is close behind Paris. London (13.1), Sydney (11.2) and Toronto (10.9) make up the rest of the top six.
- The fewest: Ghent (1.5), Antwerp (2.9), Birmingham (3.0) and Amsterdam (3.6).
- Jakarta and Bali are left out of the ranking: too few of their traffic lights are on the map to compare.
The ranking: traffic lights on a 5 km run
Average number of traffic lights crossed on a 5 km running loop starting within 3 km of the city centre. Higher means more stops.
- 1 New York 33.5
- 2 Paris 22.9
- 3 Barcelona 21.5
- 4 London 13.1
- 5 Sydney 11.2
- 6 Toronto 10.9
- 7 Warsaw 10.5
- 8 Melbourne 10.5
- 9 Berlin 9.8
- 10 Manchester 7.8
- 11 Dublin 7.6
- 12 Copenhagen 6.4
- 13 Oslo 6.2
- 14 Glasgow 5.4
- 15 Prague 5.2
- 16 Hanoi 4.6
- 17 The Hague 4.5
- 18 Edinburgh 3.8
- 19 Brno 3.7
- 20 Amsterdam 3.6
- 21 Birmingham 3.0
- 22 Antwerp 2.9
- 23 Ghent 1.5
Source: Runplanner, 1,000 generated loops, October 2026. Map data © OpenStreetMap contributors.
Full results
The average hides how much loops differ within a city, so the table also shows the middle half of the loops: half of them crossed a number of lights inside that range. No lights is the share of loops that crossed no traffic light at all. Crossings with lights is the share of all the streets and junctions a loop crossed that had a traffic light.
| # | City | Lights per 5 km | One light every | Middle half of loops | No lights | Crossings with lights |
|---|---|---|---|---|---|---|
| 1 | New YorkUnited States | 33.5 | 150 m | 27–41 | 0% | 91% |
| 2 | ParisFrance | 22.9 | 220 m | 19–26 | 0% | 63% |
| 3 | BarcelonaSpain | 21.5 | 230 m | 11–30 | 5% | 56% |
| 4 | LondonUnited Kingdom | 13.1 | 380 m | 9–18 | 0% | 34% |
| 5 | SydneyAustralia | 11.2 | 450 m | 4–17 | 10% | 49% |
| 6 | TorontoCanada | 10.9 | 460 m | 10–13 | 5% | 44% |
| 7 | WarsawPoland | 10.5 | 480 m | 7–14 | 0% | 47% |
| 8 | MelbourneAustralia | 10.5 | 480 m | 5–14 | 0% | 41% |
| 9 | BerlinGermany | 9.8 | 510 m | 8–11 | 0% | 42% |
| 10 | ManchesterUnited Kingdom | 7.8 | 640 m | 3–12 | 3% | 22% |
| 11 | DublinIreland | 7.6 | 660 m | 4–11 | 3% | 26% |
| 12 | CopenhagenDenmark | 6.4 | 780 m | 4–9 | 8% | 25% |
| 13 | OsloNorway | 6.2 | 810 m | 4–8 | 5% | 25% |
| 14 | GlasgowUnited Kingdom | 5.4 | 920 m | 2–6 | 13% | 21% |
| 15 | PragueCzechia | 5.2 | 960 m | 2–8 | 13% | 25% |
| 16 | HanoiVietnam | 4.6 | 1,100 m | 2–6 | 3% | 21% |
| 17 | The HagueNetherlands | 4.5 | 1,120 m | 2–6 | 8% | 17% |
| 18 | EdinburghUnited Kingdom | 3.8 | 1,310 m | 1–5 | 13% | 15% |
| 19 | BrnoCzechia | 3.7 | 1,350 m | 2–6 | 10% | 20% |
| 20 | AmsterdamNetherlands | 3.6 | 1,380 m | 1–6 | 13% | 13% |
| 21 | BirminghamUnited Kingdom | 3.0 | 1,690 m | 0–4 | 30% | 11% |
| 22 | AntwerpBelgium | 2.9 | 1,720 m | 1–4 | 18% | 16% |
| 23 | GhentBelgium | 1.5 | 3,240 m | 0–3 | 33% | 6% |
Download the numbers: traffic-lights-per-km-2026.csv (all 25 cities, including Jakarta and Bali).
What stands out
New York: a light at almost every corner
On the New York loops, 91% of all street crossings had a traffic light. Paris came next with 63%. Manhattan's grid puts a signalised junction at the end of nearly every block, so a 5 km run there means a light every 150 metres on average, and even the quietest of the 40 loops crossed 9.
Paris and Barcelona: a light every 220 metres
Paris (one light every 220 m) and Barcelona (230 m) come next, well ahead of London (380 m) and Berlin (510 m). Every Paris, London and Berlin loop crossed at least one light. Barcelona has more mapped traffic signals in its centre than any other city in the study, Manhattan included, but its loops differ the most: the middle half crossed anywhere from 11 to 30 lights, and two loops crossed none.
The Netherlands and Belgium: lights are rare
Amsterdam (3.6 per 5 km), The Hague (4.5), Antwerp (2.9) and Ghent (1.5) all land in the bottom half. Our guess is the separate cycle and foot paths, parks and waterside routes these cities have plenty of, which a running loop happily follows. In Ghent, one loop in three crossed no light at all.
Same country, very different runs
The five UK cities span most of the ranking: London 13.1, Manchester 7.8, Glasgow 5.4, Edinburgh 3.8 and Birmingham 3.0. In Birmingham, almost a third of the loops crossed no traffic light. Australia is closer together: Sydney 11.2, Melbourne 10.5.
Pedestrian crossings add up
Counting light-controlled pedestrian crossings as well as junction lights added 22% to the total. The effect differs a lot by city. Birmingham doubled (1.4 to 3.0 per 5 km), Barcelona went from 13.3 to 21.5 and Paris from 15.5 to 22.9, while New York, London and Sydney went up by less than 10%. Some cities simply have more crossings with their own lights; others map the light on the crossing instead of on the road, so counting only junction lights would miss it.
How we did it
- Cities. 25 cities where Runplanner users plan their runs, plus a few big running cities.
- Start points. 40 random start points per city, spread evenly within 3 km of the centre. Starts in water or anywhere without paths were skipped and replaced.
- Loops. From each start, Runplanner's loop generator drew a 5 km running loop, the same way the app does when you ask for one. It prefers paths, parks and quieter streets and does not try to avoid traffic lights. Nearly all loops came out between 4.5 and 5.5 km, so we counted lights per km and multiplied by 5.
- Lights. We counted a traffic light where the loop crosses a road at a mapped traffic light: lights at junctions and light-controlled pedestrian crossings, including UK pelican, puffin and toucan crossings. It is the same count the app shows on a route. A junction with several signals counts once. Traffic light positions come from OpenStreetMap.
- Map check. For each city we also counted the mapped junction lights in a 2 by 2 km square in the centre. Jakarta (2.0 per km²) and Bali (2.5) were far below the other cities (around 43 for the middle city), so their low counts say more about the map than the streets, and we left them out. We used junction lights for this check because cities map their pedestrian crossings in very different ways.
What this study doesn't tell you
- It counts the traffic lights in OpenStreetMap. A light that isn't on the map isn't counted, so the real numbers can be a little higher where the map is less complete.
- The loops favour paths and quiet streets. A run along main roads crosses more lights; a run that stays in one big park crosses fewer.
- 40 loops per city is a sample, not every street. The middle-half range in the table shows how much loops differ.
- It counts lights, not red lights. How long you wait depends on the timing and the time of day.
How to run with fewer traffic lights
- Start next to green. Parks, riverbanks and canal paths let you run long stretches without a crossing.
- Check before you go. The free Runplanner app for iPhone and Android shows how many traffic lights a route crosses, so you can compare two loops before you pick one.
- Let a generator try a few. The free running loop generator draws a loop of the distance you want from any start. Regenerate a few times and keep the one that suits you.
- Avoid them on purpose. Runplanner Pro, coming soon, adds an Avoid traffic lights route preference that looks for a loop where you don't have to cross at a light.
Use the data
You're welcome to use these numbers in an article, a video or your own research. Please credit "Runplanner traffic lights study (2026)" with a link to this page. The CSV has every city, including the two we left out of the ranking. Traffic light data © OpenStreetMap contributors, available under the Open Database License.
Frequently asked questions
Which city has the most traffic lights for runners?
New York, by far. Runplanner generated 40 running loops around Midtown Manhattan and they crossed 33.5 traffic lights per 5 km on average, about one every 150 metres. Paris came second with 22.9.
How many traffic lights are there on a 5K run in London?
About 13. Forty 5 km loops starting within 3 km of central London crossed 13.1 traffic lights on average, one every 380 metres. Half of the loops had between 9 and 18.
Does Amsterdam have many traffic lights for runners?
Few. Running loops around central Amsterdam crossed 3.6 traffic lights per 5 km, one every 1.4 km, and one loop in eight crossed none. The Hague came out at 4.5.
How can I avoid traffic lights on a run?
Plan routes through parks, along rivers and canals, and on quiet residential streets. The Runplanner app shows how many traffic lights a route crosses before you run it, and Runplanner Pro, coming soon, can generate loops that steer around them.
Want to see how many lights are on your own route? Open the free loop generator or plan it in the route planner, then open it in the app.