Robbery
Taking property from a person by force or threat: a mugging, a holdup at a store counter, a carjacking. How often it happens, who gets robbed, and what the records can and can't say about who does it.
Reviewed September 30, 2026
Robbery is theft with force or the threat of it. It is not burglary, which is breaking into a place, and not shoplifting or pickpocketing, which take things without force. Robbery has been one of the FBI's core counted crimes since it began collecting police reports in 1930, so it has one of the longest records of any crime.
Is robbery up?
No. Police reported 164,157 robberies in 2025, 19% fewer than in 2024. That is 51.1 for every 100,000 people, the lowest rate since 1960.
The peak came in 1991: 676,882 robberies, 286.1 per 100,000. The rate today is 18% of that.
2021 is hatched and stands apart: that year the FBI took only incident-level (NIBRS) reports, and many large agencies had not switched, so it is not joined to the years around it. Early on, fewer agencies reported, so each year's rate is per 100,000 people living where agencies reported.
Source: FBI Return A, every reporting agency's monthly count summed by CrimeDB (rules part1-return-a and part1-master); rate per 100,000 people living where agencies reported.
These are CrimeDB's own sums of every agency's monthly reports to the FBI. The FBI publishes its own national estimate, which fills in for agencies that did not report; it puts 2025 at 168,927 robberies. Both count only robberies someone reported to police.
Robberies nobody reported
When the sources are loaded, this will show (answers at: national):
- Robberies people describe in the National Crime Victimization Survey, and the share they reported to police. The survey is loaded; its robbery estimates are not built yet.
Who gets robbed?
The FBI's incident records (NIBRS) carry the detail. In 2025, agencies covering 88% of the population sent them, recording 149,397 robberies and 165,290 people robbed. Stores and other businesses are robbery victims too; the figures below count people.
67% of the people robbed were male. Most were adults of working age: 10% were under 18 and 6% were 65 or older.
| Under 18 | 10% | 15,748 | |
|---|---|---|---|
| 18 to 24 | 20% | 32,285 | |
| 25 to 34 | 25% | 40,827 | |
| 35 to 49 | 25% | 42,069 | |
| 50 to 64 | 14% | 23,740 | |
| 65 and older | 6% | 9,265 | |
| Age unknown | 1% | 1,356 |
Source: FBI NIBRS national master file, 2025, agencies that reported incident detail (CrimeDB series, rule nibrs-master).
| White | 20% | 33,595 | |
|---|---|---|---|
| Black | 23% | 37,445 | |
| Hispanic or Latino | 27% | 44,542 | |
| Asian | 6% | 9,221 | |
| American Indian, Alaska Native, or Pacific Islander | 1% | 1,935 | |
| Unknown | 23% | 38,552 |
White, Black, Asian, and the American Indian, Alaska Native, or Pacific Islander group exclude Hispanic people. Shares, not rates: a group's share also depends on how many people in it live where reporting agencies work.
Source: FBI NIBRS national master file, 2025, agencies that reported incident detail (CrimeDB series, rule nibrs-master).
Strangers are the largest group: 37% of adults robbed were robbed by a stranger. Many more have no relationship on record, often because no one was identified.
| Intimate partner, current or former | 7% | 10,511 | |
|---|---|---|---|
| Other family or household member | 2% | 2,279 | |
| Known to the victim, not family | 12% | 17,471 | |
| Stranger | 37% | 55,480 | |
| Victim was offender, or relationship unknown | 27% | 40,200 | |
| No relationship reported (unknown offender, or not collected) | 15% | 22,245 |
Source: FBI NIBRS national master file, 2025, agencies that reported incident detail (CrimeDB series, rule nibrs-master).
Who commits robbery?
Police recorded 208,791 known offenders in those robberies: anyone a victim or witness could describe, whether or not anyone was arrested. 79% were male, and 14% were under 18.
Here the record runs thin. Police recorded no race or ethnicity for 41% of known offenders, and no age for 28%. The unknowns may not be a random slice of offenders, so shares of the rest can mislead. We show them with the unknowns left in.
| White | 7% | 15,566 | |
|---|---|---|---|
| Black | 34% | 70,003 | |
| Hispanic or Latino | 17% | 34,453 | |
| Asian | 1% | 1,568 | |
| American Indian, Alaska Native, or Pacific Islander | 1% | 1,355 | |
| Unknown | 41% | 85,846 |
Known offenders are people described on a report, not people convicted. This is a count of descriptions, not a figure about any person.
Source: FBI NIBRS national master file, 2025, agencies that reported incident detail (CrimeDB series, rule nibrs-master).
| Under 18 | 14% | 28,274 | |
|---|---|---|---|
| 18 to 24 | 19% | 39,626 | |
| 25 to 34 | 20% | 41,763 | |
| 35 to 49 | 15% | 30,596 | |
| 50 to 64 | 4% | 8,588 | |
| 65 and older | 0% | 980 | |
| Age unknown | 28% | 58,964 |
Source: FBI NIBRS national master file, 2025, agencies that reported incident detail (CrimeDB series, rule nibrs-master).
Rates per 100,000 by group
When the sources are loaded, this will show (answers at: county, state, national):
- Known-offender and arrest rates per 100,000 people in each group, crude and age-adjusted.
- Victim rates for the same groups, in the same units.
Rates need the population living where reporting agencies work, by race and age. That table is not built yet, so this page shows shares. Arrests by race, a separate count from the FBI's arrest reports, are loaded and come next.
Where and when
Streets, roads, and sidewalks are the most common place, with 32% of robberies. Homes come next, with 19%.
| Highway/Road/Alley/Street/Sidewalk | 32% | 47,681 | |
|---|---|---|---|
| Residence/Home | 19% | 27,885 | |
| Parking/Drop Lot/Garage | 8% | 12,558 | |
| Convenience Store | 6% | 9,297 | |
| Other/Unknown | 4% | 6,244 | |
| Department/Discount Store | 4% | 6,190 | |
| Service/Gas Station | 4% | 5,260 | |
| Grocery/Supermarket | 3% | 5,022 | |
| All other places | 20% | 29,200 |
Location types in the FBI's own words.
Source: FBI NIBRS national master file, 2025, agencies that reported incident detail (CrimeDB series, rule nibrs-master).
Robberies are fewest around 6 a.m. and rise every hour to an evening peak around 8 p.m.. The hour starting at midnight holds more than any other, 10,755. We have not yet checked how much of that comes from reports that put an unknown time at midnight.
Each bar is the hour starting then, from 0 (midnight to 1 a.m.) to 23.
Source: FBI NIBRS national master file, 2025, agencies that reported incident detail (CrimeDB series, rule nibrs-master).
What happens in court
Police records stop at the arrest. What happens next is in court records, kept county by county. We are starting with Harris County, Texas, whose district clerk publishes its felony case files.
No data yet
When the sources are loaded, this will show (answers at: county):
- Felony robbery filings, then dismissals, pleas, and trials. Deferred adjudication shown inside pleas. Adult cases only.
- The judges who heard the cases, linked to their pages.
What the records can't answer
- Count robberies nobody reported. The survey that estimates them covers the nation only, not states or cities.
- Say who committed a robbery with no known offender. Those incidents stay their own row.
- Give group rates for offenders yet. Rates need population by group where reporting agencies work, which is not built.
- Follow a robbery from report to court. Police and court records share no case number.
How it is recorded
- 120Robbery
FBI NIBRS offense codes. The same act may carry a different name in state law and court records.
Carjacking is filed as robbery. When a robbery injures someone badly or involves a gun or knife, the summary count still files it once, as robbery; the incident record can carry the assault and the weapon too.
Where the records are
- FBI Crime Data ExplorerNational
FBI Criminal Justice Information Services (CJIS) · Incidents, Arrests, Hate crime, Use of force, Other · html, csv · annual
Official front door to UCR/NIBRS data for ~19k agencies; agency participation varies by year, especially post-2021 NIBRS transition.
Bureau of Justice Statistics · Victimization · pdf, html, csv · annual
Only national measure of crime including unreported crime; national estimates only (limited state/local detail).
US Census Bureau · Population · html, csv · annual
Denominators for rates at state/county/place/tract; ACS 1-year only for areas 65k+.
Texas Office of Court Administration · Courts · pdf, dashboard, xlsx · monthly
Aggregate caseload by county/court from CARD monthly reports; also bail and hate-crime case statistics; no names.
- Harris County District Clerk Public Datasets (criminal and civil)Local · Harris County, TX
Harris County District Clerk · Courts, Sentencing · other, xlsx · continuous
Daily/weekly/monthly .txt extracts of criminal filings, dispositions and bonds (1990-present); include defendant names; site login may be required.