Remulous Labs › UCRLens › US city crime trends
Data · FBI UCRUS city crime trends, 2018–2025
How crime has actually moved in 12 major US cities over eight years: the pandemic-era homicide spike and its steep decline, the motor-vehicle-theft wave that rose and broke, and property crime's slower drift. All figures are rates per 100,000 residents from the UCRLens API, parsed from the FBI's official Return A (RETA) master files — trends, not rankings.
How to read this data — and how not to
Do not use this page to rank cities against each other. The FBI itself cautions against ranking jurisdictions from UCR data: reporting practices, agency boundaries, and daytime populations differ in ways raw rates don't capture. Read each row across — how one city changed over time. A few honesty notes: (1) the FBI's 2021 transition to NIBRS left major gaps in the 2021 files — New York, Los Angeles, and Phoenix are effectively absent that year, and several other cities appear only partially, so 2021 is omitted from these tables and discussed separately below; (2) cells marked † reflect years where an agency's reported totals appear incomplete against adjacent years — treat them as reporting artifacts, not real drops; (3) 2025 figures come from the most recent FBI file (reta-2025.zip) and may be revised; (4) we show murder, property crime, and motor vehicle theft because they are the most consistently reported offense categories across the transition era. Every record in the API carries data_source and coverage_rate fields — verify before you rely on any figure.
The homicide spike has receded
Murder rate per 100,000 residents. Across the nine of these cities with complete reporting, murders rose from 1,898 (2018) to a peak of 2,626 (2022), then fell to 1,418 (2025) — 46% below the peak and 25% below 2018. Every city in this table is lower in 2025 than in 2022. Cities listed alphabetically.
| City | 2018 | 2020 | 2022 | 2023 | 2024 | 2025 | 2022→2025 |
|---|---|---|---|---|---|---|---|
| Atlanta, GA | 19.1 | 12.2 † | 37.5 | 29.1 | 21.0 | 20.5 | −45% |
| Chicago, IL | 20.9 | 28.6 | 22.8 | 19.0 | 17.5 | 12.4 | −46% |
| Dallas, TX | 12.0 | 19.5 | 17.6 | 20.0 | 14.5 | 11.2 | −36% |
| Denver, CO | 9.0 | 12.9 | 12.8 | 12.2 | 9.8 | 5.7 | −55% |
| Detroit, MI | 38.9 | 48.7 | 49.0 | 40.6 | 31.2 | 25.7 | −48% |
| Houston, TX | 12.2 | 18.0 | 19.5 | 15.6 | 14.1 | 11.5 | −41% |
| Los Angeles, CA | 6.4 | 8.8 | 10.2 | 8.6 | 7.0 | 5.8 | −43% |
| New York, NY | 3.5 | 5.6 | 5.3 | 4.9 | 3.9 | 3.0 | −43% |
| Philadelphia, PA | 22.1 | 31.1 | 33.4 | 26.4 | 16.9 | 14.2 | −57% |
| Phoenix, AZ | 8.0 | 10.9 | 13.2 | 11.5 | 8.4 | 7.5 | −43% |
| San Antonio, TX | 7.0 | 8.3 | 15.9 | 10.9 | 8.4 | 6.8 | −57% |
| Seattle, WA | 3.4 | 6.5 | 7.5 | 9.0 | 6.8 | 4.4 | −41% |
† Atlanta's 2020 submission appears incomplete against adjacent years. The steepest declines from the 2022 peak: Philadelphia and San Antonio (−57%), Denver (−55%), Detroit (−48%). Chicago's 2025 rate (12.4) is the lowest in its eight-year window.
The motor-vehicle-theft wave, up and back down
Motor vehicle thefts per 100,000 residents. Across the nine complete-reporting cities, the rate ran from 536 (2019) to a peak of 1,233 (2023) — a wave widely attributed to a security vulnerability in certain vehicle models — then fell 42% to 719 (2025).
| City | 2018 | 2020 | 2022 | 2023 | 2024 | 2025 | 2023→2025 |
|---|---|---|---|---|---|---|---|
| Atlanta, GA | 717 | 236 † | 728 | 1,138 | 676 | 477 | −58% |
| Chicago, IL | 372 | 373 | 799 | 1,008 | 859 | 658 | −35% |
| Dallas, TX | 756 | 824 | 1,106 | 1,511 | 1,175 | 895 | −41% |
| Denver, CO | 741 | 1,142 | 2,157 | 1,750 | 1,232 | 756 | −57% |
| Detroit, MI | 961 | 852 | 1,508 | 1,495 | 1,258 | 971 | −35% |
| Houston, TX | 521 | 650 | 784 | 892 | 718 | 532 | −40% |
| Los Angeles, CA | 430 | 529 | 669 | 676 | 260 † | 519 | −23% |
| New York, NY | 65 | 112 | 170 | 226 | 198 | 182 | −19% |
| Philadelphia, PA | 358 | 440 | 818 | 1,534 | 1,006 | 1,022 | −33% |
| Phoenix, AZ | 473 | 435 | 466 | 564 | 426 | 349 | −38% |
| San Antonio, TX | 396 | 446 | 846 | 1,281 | 836 | 614 | −52% |
| Seattle, WA | 544 | 638 | 951 | 1,227 | 974 | 732 | −40% |
† Apparent partial-year reporting; see methodology. Denver made a full round trip — 741 (2018) to 2,157 (2022) and back to 756 (2025). Philadelphia's rate more than quadrupled from 2019 (323) to its 2023 peak (1,534) and remains elevated.
Property crime: a slower drift
Property crime (burglary, larceny-theft, motor vehicle theft) per 100,000 residents. Most of these cities ended 2025 below their 2018 rate — but not all.
| City | 2018 | 2020 | 2022 | 2023 | 2024 | 2025 | 2018→2025 |
|---|---|---|---|---|---|---|---|
| Atlanta, GA | 5,004 | 1,258 † | 4,094 | 4,243 | 3,227 | 3,338 | −33% |
| Chicago, IL | 3,193 | 2,210 | 3,133 | 3,434 | 3,472 | 2,908 | −9% |
| Dallas, TX | 3,475 | 3,661 | 4,070 | 4,293 | 3,578 | 3,261 | −6% |
| Denver, CO | 3,713 | 4,661 | 6,438 | 5,765 | 4,762 | 4,103 | +11% |
| Detroit, MI | 4,318 | 3,251 | 4,487 | 4,762 | 4,305 | 3,882 | −10% |
| Houston, TX | 4,089 | 4,296 | 4,716 | 4,628 | 4,387 | 3,889 | −5% |
| Los Angeles, CA | 2,513 | 2,148 | 2,708 | 2,847 | 1,484 † | 2,234 | −11% |
| New York, NY | 1,502 | 1,558 | 2,141 | 2,426 | 2,368 | 2,241 | +49% |
| Philadelphia, PA | 3,097 | 3,006 | 4,348 | 5,073 | 4,548 | 4,305 | +39% |
| Phoenix, AZ | 3,492 | 2,990 | 2,902 | 2,484 | 2,325 | 2,027 | −42% |
| San Antonio, TX | 3,994 | 3,643 | 5,085 | 5,404 | 4,624 | 3,900 | −2% |
| Seattle, WA | 5,149 | 4,900 | 5,738 | 5,044 | 5,008 | 4,403 | −14% |
† Apparent partial-year reporting; see methodology. Phoenix has fallen steadily all eight years (−42%). New York (+49%) and Philadelphia (+39%) ended the window well above where they started, even after declining from their 2023 peaks.
The 2021 gap — why provenance matters
In 2021 the FBI retired the old Summary Reporting System and moved to NIBRS. Thousands of agencies — including some of the nation's largest — didn't complete the transition in time, so the official 2021 file simply lacks their data:
| City | What the official 2021 RETA file shows |
|---|---|
| New York, NY | Near-zero counts (murder rate 0.0, property rate 5.8) — NYPD's data is effectively absent |
| Los Angeles, CA | No usable 2021 submission |
| Phoenix, AZ | No usable 2021 submission |
A naive parser would happily report that New York had zero murders in 2021. This is exactly why every UCRLens record carries a data_source field naming the specific FBI file it came from and a coverage_rate field — so you can catch reporting artifacts instead of publishing them.
Query this data yourself
Every figure on this page comes from the UCRLens API's trends endpoint. Pull the full eight-year series for any of 17,000+ cities:
Eight-year trends for any city
curl -X GET "https://us-city-crime-statistics-fbi-ucr17.p.rapidapi.com/v1/crime/trends?city=Denver&state=CO" \
-H "X-RapidAPI-Key: YOUR_KEY_HERE" \
-H "X-RapidAPI-Host: us-city-crime-statistics-fbi-ucr17.p.rapidapi.com"See the UCRLens docs, a real sample response ↗, the API vs. raw FBI files comparison, or the OpenAPI spec ↗.