Americans reported markedly fewer violent and property crimes in 2025, with the violent victimization rate, not counting simple assault, dropping 25% from the year before, according to preliminary survey data from the U.S. Bureau of Justice Statistics.
Property crime victimization slid 19%. The figures, released the week before Tuesday's midterm elections, also capture crimes that victims never took to police.
That reach sets the National Crime Victimization Survey apart from tallies built on police records. It questions a nationally representative sample of people 12 and older about violent crimes they suffered, other than homicide, and about property crimes against their households.
By the bureau's estimate, residents in that age group endured 7.5 million violent victimizations last year, or 25.7 for every 1,000 people. Leaving out simple assault, the rate went from 9.9 per 1,000 in 2024 down to 7.4.
Aggravated assault posted the biggest drop among violent categories, 31%, sliding to 3.1 per 1,000 from 4.5. Robbery fell 27% and rape and sexual assault 11%, while simple assault barely moved, edging to 18.3 per 1,000 from 18.8.
Households absorbed an estimated 14.9 million property victimizations, a rate of 111.2 per 1,000 households, down from 138.0. Motor vehicle theft fell 30%, burglary or trespassing 25%, other household theft 19% and vandalism almost 11%.
The results track with the FBI's Uniform Crime Reporting program, which counts crimes known to police and has shown falling crime nationwide for several years. Preliminary FBI figures point to further declines in 2026.
The nonpartisan Council on Criminal Justice has found similar patterns, including drops in some of the largest U.S. cities, though trends differ from one community to the next.
A changed survey
Thaddeus Johnson, a senior fellow at the council and an associate professor of criminology and criminal justice at Georgia State University, said the agreement among data sources has strengthened his confidence that the drop in crimes other than homicide is real.
The bureau attached caveats to its numbers. It finished switching to a redesigned survey in January 2025 and says results from the new version cannot be lined up directly against those from the old one.
The revised questionnaire uses shorter, more targeted questions meant to jog respondents' memories of incidents, including unreported ones. It also digs deeper into sexual assault, how victims were treated by police and what crime cost them.
The 2025 estimates could shift when the bureau issues its final report, expected next year.
