Crash data through 2026-10-19. Updated October 24, 2026.
DC crash scorecard
DC averaged — police-reported crashes a year in —, with — people seriously injured or killed each year. The highest rate of serious injuries and deaths per square mile is in ANC —. On the streets DDOT counts traffic for (leaving out freeways), there are — crashes for every 100 million vehicle miles driven. Against 26 other US cities of similar size, DC earns a safety grade of — on traffic deaths. Click any area on the map, or pick one below, to see its scorecard.
All crashes per year, by ANC
Darker areas rank higher. Drag to move, use the buttons or pinch to zoom.
ANC —
Yearly averages, 2023 to 2025. Rank 1 means the most among ANCs.
So far in 2026 (through —): — crashes, compared with — in the same months of 2025 (—%). People seriously injured or killed: — this year, — in the same months of 2025.
Crashes by year
The lighter bar is 2026 so far. Hover a bar for its count.
Want the definitions, sources and limits behind these numbers?
What This Scorecard Shows
Traffic crashes across DC as a whole and in each of its 8 wards, 46 Advisory Neighborhood Commissions (ANCs) and 345 single-member districts, built from DC's own open data. Each area gets a safety grade, four measures with a rank against its peers (wards against wards, and so on), a trend since 2019 and its 2026 record so far. Wards and the city also show traffic deaths.
Numbers are yearly averages for —, the last three full years, so one unusual year counts less. Crash data runs through —.
How to Use It
- Show picks the level: the whole city, wards, ANCs or single-member districts.
- Color the map by shades the map by safety grade or by one measure. Darker means worse.
- Compare by sets how areas are compared: per district, in total, per square mile or per vehicle mile. It changes the tiles, ranks and table. Grades always use their own fixed recipe.
- Click an area on the map, a row in the table, or type a name in Go to (Ward 4, 4D or 4D02) to open its scorecard. The buttons under the name step down a level, from a ward to its ANCs and from an ANC to its districts.
- Click a tile to color the map by that measure.
- Show major roads draws Interstates, other freeways and principal arterials on the map. Hover a road for its name.
The Measures
Every number comes from police crash reports.
| Measure | What it counts | Shown for |
|---|---|---|
| Crashes | Every police-reported crash, whatever its severity. | All levels |
| Serious injuries and deaths | People with major injuries or who were killed: drivers, passengers, pedestrians and cyclists. The measure closest to DC's Vision Zero goal. | All levels |
| Pedestrians injured or killed | Pedestrians with any reported injury, plus pedestrian deaths. | All levels |
| Cyclists injured or killed | Cyclists with any reported injury, plus cyclist deaths. | All levels |
| Traffic deaths | Everyone killed in a crash, as the yearly average for 2019 to 2025 (seven full years, since three years has too few deaths to compare fairly), plus the total killed since 2019. | Wards and city |
Deaths are not shown for ANCs or districts because most have none in a given year, so their numbers would mostly reflect chance. Serious injuries and deaths is the better measure at those levels.
Ways to Compare
| Compare by | What it means | Best for |
|---|---|---|
| Per district | The total divided by the number of single-member districts. Each district holds about 2,000 residents (2020 Census), so this is harm per resident. For a single district it equals the total. | Fair comparisons between areas of different sizes |
| Total | The raw yearly count. | Scale; larger areas rank higher just for having more people |
| Per square mile | The total divided by land area. | Where streets and people are packed together |
| Per vehicle mile | Crashes on DDOT-counted streets per 100 million vehicle miles driven on them, leaving out freeways. | How risky the streets are once traffic, including commuters, is accounted for |
How the Grades Work
Every ward, ANC and district gets a letter for how safe it is now, plus a sign for which way it is heading.
The letter: where an area stands
Three measures, each averaged over 2023 to 2025 and compared with DC as a whole, combine into one number for how much harm an area has compared with DC:
- Serious injuries and deaths per district (40%)
- Pedestrians and cyclists injured or killed per district (40%)
- Crashes per vehicle mile, compared with the typical area at the same level (20%). Areas without enough counted traffic are graded on the other two.
| Grade | Harm compared with DC |
|---|---|
| A | Half the DC rate or less |
| B | 50% to 80% of the DC rate |
| C | Within 25% of the DC average |
| D | 1.25 to 2 times the DC rate |
| F | More than twice the DC rate |
Cutoffs are fixed rather than curved, so every area can earn an A as DC gets safer.
The sign: which way it is heading
Crashes and people hurt in the last two years (October 2024 to October 2026) are compared with the two years before. People hurt means everyone seriously injured or killed, plus pedestrians and cyclists with minor injuries.
| Sign | Meaning |
|---|---|
| Plus | Improving at least 12% faster than DC as a whole |
| No sign | Changing about as fast as DC |
| Minus | Worsening at least 12% compared with DC |
Each area is compared with DC's own change because citywide shifts, like the pandemic or a change in police reporting, happen to every area at once. The city grade is what tracks those.
Fair to small areas
Districts with only a few cases are pulled partway toward the DC average, by about three cases' worth, so one bad crash cannot sink a grade. Wards and ANCs have enough cases that this barely matters.
The City Grade
DC's own grade compares its traffic deaths per resident with the — US cities of 500,000 to 1 million people, using federal fatality data for 2018 to 2022 from a 2026 peer-reviewed study. Crash counts are not used because police departments report crashes differently, while deaths are recorded the same way everywhere.
The same letter cutoffs apply against the typical peer city. The sign compares DC's change since 2008 to 2012 with the typical peer's. The study ends in 2022 and DC's traffic deaths rose after that, so the city grade may be generous today.
Traffic and Vehicle Miles
DDOT's 2025 traffic counts give average daily traffic for about — miles of DC's busier streets, adding up to — billion vehicle miles a year. Each road segment's traffic is multiplied by its length inside each district.
- Freeways are left out entirely. Roads DDOT classifies as Interstate or Other Freeway and Expressway (I-295, I-395, the Whitehurst Freeway, the Suitland Parkway) carry — billion vehicle miles a year but have only about — crashes a year. Leaving them in made any area with a freeway look far safer than its streets are. Freeway crashes still count in every other measure, because people get hurt there.
- Which crashes count. A crash counts as on a counted street if it is recorded on that route or lies within about 30 yards of it. Citywide, —% of crashes happen on counted streets; each scorecard shows its own share under Traffic context.
- Boundary roads. Many district lines run down the middle of busy roads like Georgia Avenue. Crashes and traffic on a boundary road are split evenly between the districts on each side.
- Minimum traffic. Districts with under 2 million vehicle miles a year on counted streets get no rate, since a handful of crashes would swing it wildly. — of — districts have a rate; every ward and ANC does.
- One year of counts. The 2025 counts are applied to 2023 to 2025 crashes, assuming traffic was broadly similar across those years.
How Crashes Are Placed
- Boundaries. Each crash is placed by its reported location inside the 2023 ANC and district boundaries, the ones used in this year's elections.
- Wards are built from their single-member districts using DC's 2022 ward boundaries. Every district sits at least 97% inside one ward. The ward labels in the crash records are inconsistent in places, so they are not used.
- Edges. Crashes recorded just outside any district, usually on a boundary street, go to the nearest district within about 65 yards. Only 3 crashes since 2019 fell outside DC entirely.
- Bad records. A few hundred records with impossible dates (such as 1825 or 1900) and about 1,400 with no date were left out. 15 road segments in DDOT's traffic file share one location near the Capitol, a mapping error, and were dropped (about 0.2% of traffic).
Limits to Keep in Mind
- Police-reported crashes only. Minor crashes are undercounted everywhere. Police records can also miss someone who dies days after a crash, so death counts may run a little below DC's official totals.
- Commuter traffic. Busy roads raise an area's crash counts no matter what residents do. The per vehicle mile view helps separate road risk from traffic volume.
- Pedestrian and cyclist rates per vehicle mile measure harm against car traffic, not against how many people walk or bike there.
- Small numbers. Small districts have few crashes, so one bad year can move their rank a lot. Check the trend chart before drawing conclusions.
- Grades summarize; they do not explain. A D can come from one dangerous corridor, and the tiles below each grade show which measure drives it.
- Where, not why. Crash counts show where harm happens. Road design, speeds, enforcement and land use all play a part.
Sources
- Crashes in DC (Metropolitan Police Department crash reports), DC Open Data, opendata.dc.gov.
- Single Member District 2023 boundaries, DC Open Data.
- Ward 2022 boundaries, DC Open Data.
- Traffic Volume 2025 (average annual daily traffic), DDOT via DC Open Data.
- Roadway Functional Classification (FHWA road classes), DDOT via DC Open Data, last updated November 2022. Used to remove freeways from the vehicle-mile measure and for the major roads overlay.
- Peer city death rates: Schneider, Barbee, Nelson and Gu, "Examining a Lost Decade for Traffic Safety in Large US Cities," Findings (February 2026), Appendix Table A1, from NHTSA's Fatality Analysis Reporting System.
The crash, district and traffic files were downloaded on October 21, 2026, the ward file on October 22 and the road classification file on October 24. Please cite DC Open Data and DDOT as the sources.