Cycling Infrastructure: How to Compare Network Quality
Photo: Jan van der Wolf / Pexels
A kilometre of painted lane, a kilometre of physically separated cycle track, and a kilometre of low-traffic street are different facilities. Combining them into one total can conceal rather than reveal what a network offers its users.
Correction — 15 October 2026: We withdrew the earlier worldwide ranking because countries were assigned incompatible infrastructure categories, years, and unsupported totals. The table also contradicted the stated ranking order. Claims about a universal construction-cost ratio, fixed emissions reduction, and guaranteed participation increases were removed.
Define the inventory before ranking it
An infrastructure inventory should identify the facility types counted, the date, and the geographic boundary. Does a two-way track count once as a route or twice as directional lanes? Are recreational trails included? Are unprotected painted lanes combined with protected tracks? Does a shared path count toward both walking and cycling totals?
For a hypothetical example, City A reports 100 route-kilometres of two-way tracks. City B reports 180 lane-kilometres, counting each direction separately. The second number is larger, but the inventories do not establish that City B has more route coverage. A comparison needs compatible units first.
Quality involves more than length
England's Department for Transport provides cycle infrastructure guidance in Local Transport Note 1/20, published in 2020. It discusses route and network design using five principles: coherence, directness, safety, comfort, and attractiveness. This is a design framework rather than a worldwide country score. Official LTN 1/20 publication.
Those categories suggest practical questions for an assessment: Do the segments connect? Does the route reach useful destinations without a long detour? What happens at crossings and junctions? Is the surface and available space appropriate for the users? An inventory that answers only “how many kilometres?” leaves these questions open.
Test a whole journey
Suppose a hypothetical route has separated tracks along most of its length but requires crossing a difficult junction to reach a school. A length total counts the tracks, while a route assessment must also record the crossing. The missing connection may matter more to that journey than another kilometre added elsewhere.
This does not mean every gap has an identical safety effect. Traffic speed, volume, visibility, crossing layout, and the users' needs affect the assessment. Design guidance supplies criteria to investigate; it does not remove the need to examine the actual location.
Keep use and infrastructure separate
Cycling modal share measures travel behavior, not infrastructure length. For example, CBS reports that bicycles accounted for 27% of Dutch trips in 2023. That dated result does not tell us how many kilometres of protected track exist, establish a global infrastructure rank, or isolate the cause of participation. CBS mobility summary.
A network evaluation can report both use and infrastructure, but each needs its own source and definition. A high trip share should not silently fill a missing network inventory, and a long route inventory should not substitute for evidence of use.
Evaluate change carefully
A before-and-after count can show that recorded bicycle traffic changed on a route. To understand whether a new facility caused the change, investigate weather, season, roadworks, changes to other routes, and whether counting methods stayed consistent. A rise along one corridor may include people changing their route rather than a rise in cycling across the whole city.
Costs and benefits need equally explicit boundaries. A project estimate should state which construction and maintenance costs are included, the time horizon, and how health or travel-time benefits were modeled. There is no defensible universal return per dollar that applies to every cycling scheme.
A useful comparison checklist
Record facility category and counting unit; map connections and destinations; inspect junctions; document maintenance and accessibility; and report travel or safety outcomes with their observation periods. If a comparison combines those measures into one score, publish the weights and explain why they were chosen.
This guide does not name a worldwide infrastructure winner. It provides a way to compare documented networks without mixing incompatible totals. For the separate question of how much people actually cycle, see our modal-share guide.
Worked assessment: equal totals, different journeys
Imagine two fictional towns, each reporting 50 route-kilometres of cycle facilities. In Town A, most segments join residential areas with schools, shops, and a station. In Town B, several long segments are separated by gaps. Their inventories have equal length, but the number does not establish equal access to destinations. A map is needed to investigate the difference.
Choose a journey and trace it from origin to destination. Mark every facility transition and crossing rather than looking only at the longest uninterrupted section. Identify what category the inventory assigns to each segment. A painted lane should not be silently described as protected because an adjacent part of the route is separated.
Next make the assessment period explicit. A mapped route can be planned, under construction, or open. Counting an announced project as completed infrastructure would change the comparison's meaning. An inventory date helps readers distinguish those states instead of treating every line on a planning map as an available facility.
Finally separate coverage from condition. A length statistic cannot state whether a route's surface is maintained or whether a crossing layout meets the chosen design criteria. Those require observation or an assessment record. Avoid converting the checklist into a numerical league table unless the scoring rules and weights are published.
This is a hypothetical audit exercise, not a finding that either named country has a particular network defect. It explains how an evaluator can use the LTN design categories alongside an inventory while preserving the difference between quantity, availability, and quality. A documented comparison may report all three rather than pretending one total measures them all.
Related guides
Follow new articles
Add SciHub101 to your feed reader to see new science and technology stories.
About the Author
SciHub101 is an independent science and technology publication founded and edited by Buğra Sözeri, a software engineer and app developer based in Türkiye. AI tools assist with drafting and research. Source documentation in the existing archive is uneven, and articles are being reviewed and corrected. The Editorial Team byline identifies editorial responsibility; it does not certify specialist or independent human review. Where we compare products, anything built by our own team is labelled as such — see the editorial guidelines and About page for how corrections and disclosures work.
Related Articles
Countries Most Vulnerable to Climate Change (2025 Ranked)
Somalia, Chad, and Sudan top the Global Climate Risk Index. Small island nations like Tuvalu and Kiribati face complete disappearance beneath rising seas. Bangladesh could lose 17% of its land by 2050. Here's which countries face the greatest climate threat — and the cruel injustice that those most at risk contributed almost nothing to causing the crisis.
Countries with the Most Desert Land (2025 Ranked)
Russia has the most desert and dryland by total area. But Australia has the highest percentage of desert territory at 35%. The Sahara covers 11 countries. Desertification is advancing at 12 million hectares per year — an area the size of England lost annually to spreading desert. Here's the full ranking and the crisis of expanding deserts.
Cycling by Country: How to Interpret Trips and Modal Share
How to compare cycling trip shares, commute statistics, and national travel surveys without confusing cities with countries.
Countries with the Most Electric Vehicles (Ranked)
China dominates global EV adoption: of the nearly 58 million electric cars on the world's roads at the end of 2024, China accounts for more than half. Chinese buyers purchased over 11 million new EVs in 2024 alone — more than were sold globally just two years earlier. Norway leads in EV penetration, with 29% of all cars on its roads now electric. This ranking covers the top countries by total EV stock, new sales, and market share, sourced from the IEA Global EV Outlook 2025.
Countries with the Most Reliable Electric Grids (Regulator Data)
Germany's regulator reported 12.8 minutes of unplanned outage per customer in 2023. US customers averaged about 11 hours in 2024, a hurricane year, and about 2 hours in routine conditions. A clean global ranking does not exist because countries count outages differently, so this article compares what each national source actually reports.
Smart Cities and AI — How Machine Learning Reduces Urban Energy Waste
Cities use roughly two-thirds of the world's energy, and much of it runs on fixed schedules. Machine learning can forecast demand and adjust traffic signals, lighting, and building controls ahead of time. Here is what has been measured, what is vendor claim, and where the method breaks.