WHAT THE DATA SHOWS
A written analysis of cyclist fatality figures across 39 countries, drawn directly from the Road Safety Data Explorer, with the methodology and sources behind every number.
Cyclist fatalities: how many, and where
Road safety debates stall in a predictable way: everyone argues from different numbers, and the discussion becomes about whose figures to trust instead of what to do. The explorer exists so that researchers, journalists, advocates, policymakers, and our own members can start from the same figures. Here is what they show.
The 30 countries in the explorer that report a full road user breakdown recorded 3,219 cyclist fatalities in their most recent reporting year, combined. For 27 of those countries that year is 2023; Argentina and Australia report 2022, and Norway reports 2024. The three highest counts in the dataset are Japan with 500 cyclist fatalities in 2023, Germany with 446, and Colombia with 435. The United States, reported separately through the Fatality Analysis Reporting System rather than the international matrix, recorded 1,166 cyclist fatalities in 2023.
Where cycling is common, cyclists account for a striking share of all road fatalities. In the Netherlands, 34.2% of 2023 road fatalities were cyclists. Belgium recorded 20.2% and Switzerland 17.4% in the same year, while Greece recorded 2.0% and Canada 2.3%. That share tracks how much a country cycles at least as much as how dangerous its roads are, and that is precisely why it matters: as cycling grows, the exposed population grows with it.
The ten-year trend: falling behind overall progress
The broader road safety story is one of progress. Total road fatalities fell between 2014 and 2024 in 25 of the 38 countries with data for both years. Cyclists are not sharing equally in that progress. In France, cyclist fatalities rose 41% between 2014 and 2024 while total road fatalities fell 5.6%. In Germany, cyclist fatalities rose 12% against an 18% fall overall. Switzerland recorded a 33% rise for cyclists against a 2.9% rise overall, and Spain a 22% rise against 5.7%. In the United States, cyclist fatalities climbed from 729 in 2014 to 1,166 in 2023, an increase of about 60%, while total road fatalities rose 25% over the same period.
The trend is not universal, and that is the most important thing the data shows. Over the same 2014 to 2024 period, cyclist fatalities fell 63% in Lithuania, 61.5% in Slovenia, 59.2% in Chile, 53% in Hungary, and 50% in Norway. Improvement for cyclists is achievable. It is simply not automatic.
Rates, e-bikes, and what these numbers do not capture
Measured against population, the safest roads in the dataset in 2024 were in Norway, with 1.6 road fatalities per 100,000 people, Sweden with 2.0, and Denmark with 2.4. The highest rates were in Costa Rica with 15.4, Colombia with 15.3, and Uruguay with 12.8. The United States recorded 12.21 per 100,000 in 2023. These are all-road-user rates: the dataset deliberately does not offer a cyclist rate per 100,000 inhabitants, because a meaningful one would need exposure data, how much cycling actually happens, that most countries do not collect.
The composition of cyclist fatalities is also changing. In Germany, 44% of cyclist fatalities in 2024 involved an e-bike. In Switzerland the figure was 55%. As electric bicycles broaden who cycles, how far, and how fast, they are reshaping who is at risk on the road.
Finally, these figures understate the problem, and it is worth being clear about how. Serious injuries are recorded far less reliably than fatalities, particularly for crashes that never involve a motor vehicle: the US census, for instance, counts only cyclist fatalities from crashes involving a motor vehicle in transport. Definitions differ between countries and over time, and the US figures include e-bike riders only from 2022 onward. Even within a single country the numbers can disagree: the Netherlands' 2023 road user matrix sums to 608 fatalities while its published headline count for the same year is 684, a gap the source report does not reconcile. None of these figures adjust for exposure. The data is the best available; it is not the whole picture.
Methodology and sources
International figures come from the International Transport Forum road safety database, as published in the IRTAD Road Safety Annual Report 2025 (OECD/ITF, published February 2026, data as reported up to 12 December 2025). United States figures come from the National Highway Traffic Safety Administration's Fatality Analysis Reporting System, 2023 data. FARS is a census of fatal crashes rather than a sample: it records every fatality occurring within 30 days of a crash on a public trafficway that involved a motor vehicle.
Rates per 100,000 inhabitants divide a country's road fatalities by its population, using United Nations World Population Prospects 2024 estimates. Population rates make countries of different sizes comparable, but they carry a limitation worth naming: they do not account for how much cycling actually happens in each country. A country where few people cycle can post a low cyclist toll and still be a dangerous place to ride.
Comparability caveats apply throughout. Countries differ in reporting years, in how they define a road fatality, and in how completely they report the breakdown by road user. The US FARS figures are not directly comparable to the international matrix, and vehicle classification changes mean some US vehicle-type figures are not comparable across 2019 and 2020. Where a figure is not available in the source, the explorer shows that plainly rather than estimating.
Last reviewed: 5 August 2026. The underlying dataset is free to reuse. If you build on it, please cite the original sources: ITF (2025), Road Safety Annual Report 2025, OECD Publishing, Paris; and NHTSA, Fatality Analysis Reporting System, 2023.
Check the Numbers Yourself
Every figure on this page can be filtered, compared, and downloaded in the explorer, at no cost and with no registration.
