How the scorecard is built.
The INRIX Global Traffic Scorecard compares congestion across — urban areas in — countries. This page defines every figure in the report — hours lost, the ranks, speeds, change, cost per driver and the drive-time maps — and states, at the same volume, what has no reading.
What the scorecard compares#
The scorecard compares — urban areas in — countries, across — regions. Every figure describes an urban area: a city and the built-up area around it, as INRIX defines it for the scorecard. Urban areas are compared with each other. A country's figures on the ranking page are the median of its own urban areas, not a separate national reading.
The scorecard reports — years:
The Year control on the ranking and city pages switches every table and figure between them. The overview is pinned to —.
Hours lost#
Hours lost is the time a driver loses to congestion over a year of peak-hour commuting, against the same trips driven at free-flow speed. It is the headline figure, and every rank is built on it.
It is calculated from INRIX speeds, week by week, for each urban area:
- The commute. A typical trip in the urban area — its mean commute length — is timed inbound at the morning-peak speed and back out at the evening-peak speed.
- Free flow. The same two trips are timed at the reference overnight speed, taken from 2023.
- The week's delay. The difference between the two, over five commuting days. Morning and evening together are the week's total.
- The year. 2025 hours lost are the annual baseline carried in the delivery. 2026 hours lost carry that baseline forward by the year-on-year change in total weekly delay: the weeks of 2026 the data covers, against the same weeks of 2025.
2026 hours lost = 2025 hours lost × (1 + year-on-year change)
In — the median urban area lost —, and the figures run from — to — in —. The median is unweighted: every urban area counts once.
The ranks#
Each urban area carries three ranks in each year. Rank 1 is the most congested urban area in every year, on every rank.
—
Impact rank
The impact rank is the scorecard's official rank. It multiplies hours lost by the natural log of the urban area's population, so that a large urban area ranks above a small one with the same delay, while population weights the hours rather than replacing them. It is the rank the ranking page orders by.
A rank is printed out of the urban areas ranked in that year's edition:
Hours rank
The hours rank orders urban areas by hours lost alone, the most hours ranked 1. It leaves population out, so it can differ from the impact rank.
Country rank
The country rank is the impact rank within the urban area's own country. In — it is out of the number of urban areas the country has in the scorecard. In earlier years it is out of the highest country rank delivered for that country, because those ranks were set against each edition's own, larger set of urban areas.
Speeds#
Three speeds describe the same commute. Each is published in miles per hour and shown in the reader's units — the Units control switches between mph and km/h.
- Peak speed — the commute's typical speed in the morning and evening peaks.
- Off-peak speed — the same trip between the peaks.
- Last-mile speed — the morning-peak speed over the final mile of the trip in.
Across the urban areas in —, the median peak speed is —, the median off-peak speed — and the median last-mile speed —.
Speeds for 2024 are not like-for-like. A change in speed between 2024 and a later year says as much about the model as about the road.
Change#
Year-on-year change is the change in total weekly delay over the weeks of 2026 the data covers, against the same weeks of 2025. It is the figure that carries 2025's hours lost forward to 2026.
Change vs 2024 compares 2026 hours lost with 2024 hours lost: (2026 − 2024) ÷ 2024. It is measured against 2024, and only 2024 — it is not a comparison with any earlier baseline.
Both are printed as a signed percentage. A rise means more hours lost.
Cost per driver#
Cost per driver is the annual cost of the time a driver loses to congestion: 2026 hours lost multiplied by the hourly cost of a driver's time in that country. The hourly costs are the 2025 values, rolled forward by inflation. It is published for 2026 only.
It is published in the country's own currency, and for — countries only:
That is — urban areas. The other — print No reading: there is no hourly cost for their countries in this edition, and the scorecard does not borrow one.
The exchange rate#
The Prices control shows each cost in its local currency by default, or converts every cost to US dollars, pounds sterling or euros, so urban areas in different countries can be compared on one scale. Conversions use the European Central Bank's euro reference rates of —, pinned for the edition. A converted cost is that one day's rate applied to the published figure.
- Rate as of
- —
- Euro to US dollar
- —
- Euro to pound sterling
- —
Working days#
Hours lost are also stated in working days: hours lost divided by —, —. It is the only comparison the scorecard draws. The median — in — is —.
Drive-time maps#
Each city page maps how far a driver gets in — minutes from the — — the centre of the urban area's outline, not a landmark — at every —-minute step through the day, a —-step day. The years — are drawn over each other, so the reach can be compared by time of day and from year to year.
— urban areas have no drive-time reading and no map location: —. They are ranked like every other urban area.
Busiest corridors#
A separate — study names the busiest corridors in each urban area it covers: stretches of road, each delivered as a line on the map with its travel time at every hour of an average day and in free-flowing traffic. It covers urban areas in — only, and the — edition only, so the corridor figures appear only when that year is selected.
Hours of delay = (travel time at the worst hour − free-flowing travel time) × — working days ÷ 60
The worst hour is the hour of the average day with the longest travel time. Hours of delay is the time a driver would lose in a year by driving the corridor at that hour on every working day. It describes one corridor, not a driver's whole commute, and it is a different figure from hours lost: the two are never compared or added.
The study keeps the top 30 corridors in most urban areas and the top 40 in a capital, so a large urban area can lose a few of its lowest-ranked corridors. The site publishes — corridors across — urban areas, numbered 1, 2, 3… within each urban area by hours of delay. The delivery names no roads, so neither does the site. Road class is the functional road class as delivered, 1 being the highest. The median corridor carries — hours of delay; the most is — hours.
—
What has no reading#
A figure with no reading prints No reading, never a zero and never an estimate. Ranks are out of the urban areas ranked, and a table sorts blanks last in either direction. In this edition:
- — US metro areas are absent from the delivery and so from the scorecard: —. They are not ranked, and no figure on the site describes them.
- Cost per driver exists only for —. Everywhere else it prints No reading.
- Drive-time maps and map locations are missing for —.
- Busiest corridors exist only for —, and only for —. Every other urban area says so on its page; under any other year the corridor section is not shown.
- Population. — It enters the impact and country ranks only.
- Outliers. The outlier review for this edition has not been published. Every figure is printed as delivered.
- Earlier years' ranks are each edition's own, so the numbering has gaps among this edition's urban areas — see The ranks.
Sources#
Speeds, delays and the drive-time maps come from INRIX traffic data. Every figure on the site is read from the 2026 delivery:
The 2025 and 2024 figures are this delivery's own. Its 2024 is the 2024 data as restated in the 2025 edition. Exchange rates are the European Central Bank's. Population is matched to each urban area upstream.
Data files#
The site reads only the files below. Each is published as JSON beside the pages.