Crime statistics by race: FBI arrests by race, sex and age
Of the 7,138,899 arrests reported to the FBI in 2025, 63.0% were of people the arresting officer classified as White and 30.5% as Black or African American.
An arrest is an event, not a crime or a person: one person arrested three times is three arrests, and an arrest is not a conviction. Race is the arresting officer’s classification, not what the person arrested says about themselves. The FBI publishes no Hispanic or Latino field with arrests, so nothing here is by ethnicity; the full caveats are below the tables.
By race · 2025
2 more columns on a wider screen
| Race | Arrests | Share of arrests |
|---|---|---|
| White | 4,499,167 | 63.0% |
| Black or African American | 2,180,417 | 30.5% |
| Unknown | 158,164 | 2.2% |
| American Indian or Alaska Native | 149,669 | 2.1% |
| Asian | 114,458 | 1.6% |
| Native Hawaiian or Other Pacific Islander | 37,024 | 0.5% |
The rate is arrests per 100,000 residents of that group, using Census Bureau population estimates for the same year. The categories are the FBI’s own, including its legacy combined grouping “Asian, Native Hawaiian, or Other Pacific Islander”, which it publishes and leaves empty.
The FBI publishes 3 further race categories — “Asian, Native Hawaiian, or Other Pacific Islander”, “Multiple” and “Not Specified” — and has returned zero in every one of the 260 scope-years published since 2022: the nation and every state, every year, without a single value between them. They are named here rather than shown as zeros, because a zero in a column of numbers is read as a measurement and this is an empty category.
344 of 7,139,243 arrests (< 0.1%) carry no race category at all, so the rows above total slightly less than the arrests recorded, and every share is of the rows’ total.
Arrests per 100,000 residents ranged from 484 to 4,715 across the FBI's race categories in 2025
Arrests reported to the FBI per 100,000 residents of each group, using Census population estimates for the same year. An arrest is an event, not a person.
A rate compares arrests with the size of a group. It assumes nothing about offending, and the data cannot say why the rates differ: where officers were, whom they stopped, what was reported and what happened are not separated in it.
The denominator is each race alone across all origins, because the FBI records Hispanic arrestees under a race and publishes no ethnicity field.
Groups with fewer than 20 arrests, or with no Census counterpart, have no bar. Their counts are in the table above.
Source: The FBI’s Crime Data Explorer
About this chart
The figures
| Race, as the arresting agency recorded it | Arrests per 100,000 residents |
|---|---|
| Black or African American | 4,715 |
| Native Hawaiian or Other Pacific Islander | 3,880 |
| American Indian or Alaska Native | 2,124 |
| White | 1,820 |
| Asian | 484 |
Embed this chart
Paste this where you want the chart. The picture carries its title, source and credit, under the site’s license.
Homicide victims and offenders by race · 2025
What race are murder offenders in the US?
Of the 17,242 homicide offenders police described to the FBI for 2025, 45.6% were recorded as Black or African American and 34.7% as White, and for 17.4% no race was recorded. An offender here is the person police identified, not someone convicted, and a killing with no one identified has no offender in these counts.
The reports count victims and offenders separately, so they cannot say how often a killing crossed racial lines, and an offender’s race is known only where the police identified one. Race and Hispanic or Latino origin are recorded separately. Nothing here says why the shares differ.
Homicide only. These are the FBI’s Supplementary Homicide Reports, which describe murder and non-negligent manslaughter and nothing else. They cover 95.4% of the 14,380 homicides the FBI counted in 2025, so they describe most homicides and not all of them.
Victims and offenders are two counts, not pairs. A victim’s race is not matched to the offender’s, so this table cannot say who was killed by whom, and nothing here is a rate: there is no population beside each race.
How much race is not recorded. The FBI’s “Unknown” and “Not Specified” are kept in the table, last. Together they are 17.4% of offenders (2,998 of 17,242) and 2.0% of victims (268 of 13,714). An offender is only known where one was identified, so the offender column describes solved homicides more than all of them.
2 more columns on a wider screen
| Race | Victims | Share of victims |
|---|---|---|
| Black or African American | 6,928 | 50.5% |
| White | 6,094 | 44.4% |
| Asian | 212 | 1.5% |
| American Indian or Alaska Native | 178 | 1.3% |
| Native Hawaiian or Other Pacific Islander | 34 | 0.2% |
| Unknown | 268 | 2.0% |
| Not Specified | not reported | not reported |
Each share is of that column’s own total. A race label the FBI publishes and has never filled is not printed as a zero: for victims, Asian, Native Hawaiian, or Other Pacific Islander, Multiple, Not Specified; for offenders, Asian, Native Hawaiian, or Other Pacific Islander, Multiple. The full breakdowns by relationship, weapon, circumstance, age and sex are on the homicide victims and circumstances page.
By sex · 2025
- 5,186,051
- Male
- 3,092 per 100,000 residents
- 1,953,192
- Female
- 1,122 per 100,000 residents
By age · 2025
The FBI publishes age separately for each sex, in its own bands: single years from 15 to 24, wider bands on either side. Census single years of age sum into those bands exactly, which is why a rate is possible here at all.
1 more column on a wider screen
| Age | Sex | Arrests |
|---|---|---|
| Under 10 | Female | 299 |
| Under 10 | Male | 1,225 |
| Under 11 | Female | 0 |
| Under 11 | Male | 0 |
| 10-12 | Female | 11,546 |
| 10-12 | Male | 22,671 |
| 11-12 | Female | 0 |
| 11-12 | Male | 0 |
| 13-14 | Female | 37,574 |
| 13-14 | Male | 72,578 |
| 15 | Female | 27,916 |
| 15 | Male | 60,889 |
| 16 | Female | 29,884 |
| 16 | Male | 71,676 |
| 17 | Female | 30,014 |
| 17 | Male | 80,708 |
| 18 | Female | 38,604 |
| 18 | Male | 106,621 |
| 19 | Female | 40,775 |
| 19 | Male | 112,115 |
| 20 | Female | 41,391 |
| 20 | Male | 112,174 |
| 21 | Female | 43,344 |
| 21 | Male | 117,615 |
| 22 | Female | 44,100 |
| 22 | Male | 118,400 |
| 23 | Female | 45,832 |
| 23 | Male | 120,391 |
| 24 | Female | 48,745 |
| 24 | Male | 125,021 |
| 25-29 | Female | 256,139 |
| 25-29 | Male | 666,069 |
| 30-34 | Female | 294,928 |
| 30-34 | Male | 770,964 |
| 35-39 | Female | 290,020 |
| 35-39 | Male | 724,493 |
| 40-44 | Female | 236,025 |
| 40-44 | Male | 599,245 |
| 45-49 | Female | 160,733 |
| 45-49 | Male | 433,861 |
| 50-54 | Female | 104,034 |
| 50-54 | Male | 301,324 |
| 55-59 | Female | 75,133 |
| 55-59 | Male | 238,384 |
| 60-64 | Female | 51,691 |
| 60-64 | Male | 178,397 |
| 65 and over | Female | 44,465 |
| 65 and over | Male | 151,230 |
| Juvenile Other | Female | 0 |
| Juvenile Other | Male | 0 |
| Adult Other | Female | 0 |
| Adult Other | Male | 0 |
Over time · 2022–2025
The share of arrests each category accounted for, in the 4 complete years the FBI publishes since agencies finished moving to incident-level reporting. Shares rather than counts or rates, because the FBI publishes no participating-agency figure with arrest data, and a count that rises because more agencies reported looks exactly like a count that rises because more people were arrested.
3 more columns on a wider screen
| Race | 2022 | 2023 |
|---|---|---|
| White | 66.4% | 64.2% |
| Black or African American | 27.6% | 28.8% |
| Unknown | 2.0% | 3.0% |
| American Indian or Alaska Native | 2.2% | 2.1% |
| Asian | 1.3% | 1.4% |
| Native Hawaiian or Other Pacific Islander | 0.5% | 0.4% |
| Multiple | 0.0% | 0.0% |
| Not Specified | 0.0% | 0.0% |
How much of this is a change in who reported. It can be measured, because the FBI publishes the same figures by state and states move independently. National arrests rose 8.3% between 2022 and 2023. In the same year New York rose 121.8% — 204,565 arrests from one jurisdiction, against 539,374 nationally — while the median state moved 5.4%. That is agencies entering the reporting system, not one state arresting more than twice as many people in a year.
Recomputing the national shares with New York left out moves the four-year change in the largest category from −3.4 points to −2.4 points. So roughly a third of the movement in this table is one state joining, and the rest survives removing the single largest such event. Both halves are stated because neither on its own is the answer.
By offense · 2025
The FBI groups arrests by offense in three overlapping vocabularies. This is one of them — its offense categories — and the three are never added together. No rate: there is no population a denominator could express as being at risk of an offense.
These are not shares of the arrests above. The offense rows come to 7,203,315 against the 7,139,243 arrests recorded in 2025 — 0.9% more, and more in earlier years. So an offense divided by the arrest count on this page is not that offense’s share of arrests. Why the two differ is not something the FBI publishes, and it is not guessed at here.
Not by race. The FBI serves arrests by offense and arrests by race as separate totals, and its arrest endpoint refuses any offense but “all”, so no offense can be split by race here.
| Offense | Arrests |
|---|---|
| All Other Offenses | 2,442,314 |
| Simple Assault | 978,293 |
| Drug/Narcotic Offenses | 860,265 |
| Drive Under the Influence | 737,181 |
| Larceny | 661,814 |
| Aggravated Assault | 309,508 |
| Disorderly Conduct | 257,493 |
| Weapon Law Violations | 152,513 |
| Destruction/Damage/Vandalism of Property | 151,046 |
| Burglary | 101,804 |
| Liquor Law Violations | 94,747 |
| Fraud Offenses | 85,523 |
| Robbery | 60,690 |
| Stolen Property Offenses | 57,906 |
| Motor Vehicle Theft | 53,968 |
| Family Offenses, Nonviolent | 42,565 |
| Vagrancy/Loitering | 39,197 |
| Counterfeiting/Forgery | 31,690 |
| Sex Offenses, Non-forcible | 18,295 |
| Sex Offenses | 17,789 |
By state · 2025
Arrests per 100,000 residents of each group, in each of the 51 jurisdictions that reported, with the same floor and denominators as the national table. A dash is a group with fewer than 20 arrests or no Census counterpart. Each state’s page has its race, sex and age tables in full.
3 more columns on a wider screen
| State | White | Black or African American |
|---|---|---|
| Alabama | 2,234 | 4,792 |
| Alaska | 2,086 | 6,131 |
| Arizona | 2,479 | 6,907 |
| Arkansas | 2,708 | 7,549 |
| California | 2,115 | 4,195 |
| Colorado | 3,162 | 8,594 |
| Connecticut | 1,843 | 4,825 |
| Delaware | 1,879 | 5,412 |
| District of Columbia | 1,897 | 14,845 |
| Florida | 1,101 | 2,909 |
| Georgia | 1,211 | 3,017 |
| Hawaii | 4,143 | 9,776 |
| Idaho | 2,203 | 7,022 |
| Illinois | 1,203 | 4,360 |
| Indiana | 1,547 | 5,759 |
| Iowa | 2,034 | 12,415 |
| Kansas | 2,108 | 9,248 |
| Kentucky | 3,279 | 8,386 |
| Louisiana | 1,275 | 3,046 |
| Maine | 2,065 | 6,183 |
| Maryland | 1,411 | 3,168 |
| Massachusetts | 1,257 | 2,967 |
| Michigan | 1,316 | 4,855 |
| Minnesota | 1,193 | 6,664 |
| Mississippi | 1,686 | 3,735 |
| Missouri | 1,751 | 6,375 |
| Montana | 1,814 | 10,212 |
| Nebraska | 2,366 | 11,414 |
| Nevada | 3,291 | 9,992 |
| New Hampshire | 2,532 | 8,489 |
| New Jersey | 1,365 | 4,211 |
| New Mexico | 2,311 | 5,186 |
| New York | 1,374 | 5,223 |
| North Carolina | 1,607 | 4,634 |
| North Dakota | 2,935 | 10,051 |
| Ohio | 1,082 | 4,081 |
| Oklahoma | 1,969 | 6,707 |
| Oregon | 3,054 | 9,670 |
| Pennsylvania | 1,589 | 5,094 |
| Rhode Island | 1,830 | 5,725 |
| South Carolina | 2,138 | 4,969 |
| South Dakota | 2,135 | 11,660 |
| Tennessee | 3,036 | 6,973 |
| Texas | 1,709 | 3,998 |
| Utah | 2,942 | 13,691 |
| Vermont | 2,176 | 10,972 |
| Virginia | 1,945 | 4,716 |
| Washington | 1,889 | 5,789 |
| West Virginia | 1,598 | 5,710 |
| Wisconsin | 2,507 | 12,004 |
| Wyoming | 2,765 | 11,171 |
What these figures are and are not
- An arrest is an event, not a person. One person arrested three times in a year is three arrests, so these are not counts of individuals and a rate is not the share of a population that was arrested.
- Race is recorded by the arresting agency, not reported by the person arrested.
- The FBI publishes no ethnicity field with these arrests, so no Hispanic or Latino figure can be derived from them. Hispanic arrestees are recorded under a race, which is why the population each rate divides by is that race across all origins rather than the non-Hispanic part of it.
- The FBI’s “Multiple” category is an arresting officer’s classification of one person. The Census Bureau’s “two or more races” is what a person says about themselves. They are not the same population, so that row carries a count and no rate.
- Reporting to the FBI is voluntary, and the FBI publishes no participating-agency figure with arrest data, so a change in an arrest count cannot be separated from a change in who reported. The size of that is measurable from its own state figures: national arrests rose 8.3% between 2022 and 2023, while NY alone rose 121.8% — 204,565 arrests from one state, against 539,374 nationally. Counts are not comparable across that year.
- A rate is not published where fewer than 20 arrests lie behind it. That is the convention US federal vital statistics use for an unreliable rate, and it matters most in the smallest groups: nineteen arrests among a few hundred residents produces a figure that one or two arrests would move by a third. The count is always shown.
- Comparing the composition of arrestees with the composition of a population assumes equal offending and equal likelihood of arrest across groups. This data supports neither assumption, and nothing here implies either.
- 344 arrests in 2025 carry no race at all — not even “unknown” — so the race rows total slightly less than the arrests recorded.
Why this is arrests, and not crimes
There is no national count of crimes committed by race, and no US federal source publishes one. What the FBI publishes is arrests: an event in which an agency took someone into custody and recorded a race for them. An arrest is not a conviction, most offenses never produce one, and the offenses that do are not a random sample of the offenses that happen.
So the tables below say who was arrested, as recorded by the arresting agency. They do not say who offended, and the difference between those two is not a rounding error — it is everything this page cannot tell you.
Offenders and victims for other offenses
For every other offense, arrests are the only breakdown by race the FBI serves. Its incident-level endpoint advertises offender and victim demographics and answers every request with the complete structure and no values in it, checked across 2019, 2021, 2022, 2023 and 2024. So there is no offender breakdown here for other offenses, and no figure anywhere on this site is Hispanic or Latino: the arrest data has no ethnicity field at all.
On the same question
- Arrests by race and sex, and what the data cannot support — the analysis: what the denominators are, and why
- What percentage of crimes are committed by each race? — the question, answered in a sentence
- Homicide victims and circumstances — relationship, weapon and circumstance, by race and age
Elsewhere in this hub
Sources
- FBI Crime Data Explorer — Arrests — Federal Bureau of Investigation, Criminal Justice Information Services Division · Public domain — U.S. Government work · retrieved September 19, 2026
- Census Population Estimates — state characteristics — U.S. Census Bureau
Compiled automatically from FBI Crime Data Explorer — Arrests and Census Population Estimates — state characteristics by a checked SQL query · Editor: Kian O'Connor · How this site is built
Data retrieved