Serious violent crime in the United States fell by a quarter in 2025, according to preliminary estimates the Bureau of Justice Statistics released Monday from its annual survey of crime victims. The rate of violent victimization excluding simple assault dropped from 9.9 per 1,000 residents age 12 or older in 2024 to 7.4 per 1,000 last year.
Property crime against households fell almost as sharply. The rate of property victimization declined 19 percent, from 138.0 per 1,000 households to 111.2, and the bureau estimated that households experienced 14.9 million property victimizations over the year.
The Justice Department, in its announcement of the figures, said the aggravated assault rate fell 31 percent, from 4.5 to 3.1 per 1,000 people. Burglary or trespassing was down 25 percent, motor vehicle theft 30 percent and other household theft 19 percent. The department described the declines in violent crime, property victimization, burglary or trespassing, motor vehicle theft and other household theft as significant.
The bureau's measure of serious violence covers rape or sexual assault, robbery and aggravated assault, and it leaves out simple assault. Counting every form of violent victimization the survey tracks, the bureau put the 2025 total at 7.5 million, a rate of 25.7 per 1,000 residents age 12 or older.
Associate Attorney General Stanley Woodward credited the administration. "This report is proof that President Trump's 'Make America Safe Again' agenda is working," he said in the department's release. "Under President Trump and Attorney General Blanche, the safety of law-abiding Americans will always come first."
The survey covers only nonfatal crimes. The department noted that a separate bureau report published last week found homicide rates fell 17 percent in 2025 from the year before.
Unlike police tallies, the survey asks a nationally representative sample of residents about crimes they experienced, whether or not they reported them. The department said the bureau gathers that information each year from about 240,000 people age 12 or older in about 150,000 households, and it describes the survey as the nation's primary source of information on criminal victimization.
The new numbers carry caveats that the bureau spells out. They are preliminary and could change when it publishes final estimates in its Criminal Victimization, 2025 report, expected in 2027. They are also the first estimates produced entirely with a redesigned questionnaire, which the bureau adopted fully in January 2025, and the bureau says results from the new version are not comparable to those from the legacy survey.
That warning matters for anyone lining the 2025 figures up against older years. The bureau phased the redesign in during 2024 with a split sample, running the old and new instruments side by side, and its comparison of the two found the new questionnaire generally produced larger estimates of both violent and property victimization. The redesigned version recorded a 2024 violent victimization rate of 28.7 per 1,000, 23 percent higher than the 23.3 per 1,000 measured by the legacy instrument.
Police data compiled outside the federal government point in the same direction. The Council on Criminal Justice, in a review of reported crime in 40 large cities, found 11 of 13 offense types lower in 2025 than in 2024. Homicides fell 21 percent in the 35 cities that reported them, or 922 fewer killings, while robberies dropped 23 percent and motor vehicle thefts 27 percent. Drug offenses rose 7 percent, and sexual assaults held even.
The council's researchers were more guarded about the reasons. They wrote that the declines are likely the result of "a complex tangle of broad social and technological changes and direct policy interventions," and that their report is not evidence of any policy's success or failure. Some of the improvement also predates last year: motor vehicle thefts in the study cities had climbed from the summer of 2020 through 2023 before turning down in 2024, and homicides there ended 2025 at 25 percent below their 2019 level. The group cautions that its city sample is not necessarily representative of the nation as a whole and that its figures are preliminary.