Findings
Using the FBI 2025 Supplementary Homicide Report master file retrieved October 7, 2026:
| Group | Recorded offender entries | July 1, 2025 population | Per 100,000 |
|---|---|---|---|
| Black female, any ethnicity | 769 | 24,172,990 | 3.1812 |
| White male, any ethnicity | 4,896 | 122,145,300 | 4.0083 |
White males have a 26.0% higher recorded-offender rate; Black females have a 20.6% lower rate. These are all-age crude recorded-offender rates, not nationally weighted estimates of all actual offending or conviction rates. Children are included in both denominators.
Non-Hispanic White comparison
Among White-male offender entries: 1,697 are coded non-Hispanic, 1,807 Hispanic, and 1,392 have unknown or blank ethnicity. The non-Hispanic White-male population is 94,527,292. Assigning none versus all of those 1,392 missing-ethnicity White males to non-Hispanic yields 1.7952 to 3.2678 per 100,000. This is a sensitivity range for ethnicity assignment only, not a confidence interval and not a bound on true national offending. It straddles the Black-female rate of 3.1812.
If missing White-male ethnicity is allocated in the same proportions as known White-male ethnicity, the resulting non-Hispanic White-male rate is 2.5084, and Black females are 26.8% higher. This is an assumption-based illustration, not an FBI estimate or an established result. Missingness may be systematic.
Counting method and limits
Read fixed-length 268-character SHR records. Exclude geographic division 0 (territories), homicide code other than A (murder/nonnegligent manslaughter), and first-offender circumstance 80 or 81 (justifiable homicide). Count first offender and all additional offender slots, up to the format capacity of 11 offenders per incident. Sex/race positions are 83/84 for first offender; additional 12-character offender blocks start at position 149. Ethnicity is position 85. All positions are one-based.
12,406 retained incidents contain 13,375 victims and 16,632 available offender entries. Some entries represent unknown offenders. 11 additional offender entries exceed the fixed-format slot capacity and cannot be assigned race/sex from this file. Unknown, blank, and not-specified race/sex are excluded from group numerators; no imputation or national coverage adjustment is applied. Offenders can recur across incidents; these are not deduplicated unique individuals. Multiple offenders can be linked to a single killing, and one offender to multiple victims. The result measures recorded offender involvement, not an exact count of killings uniquely attributable to each group.
The FBI published annual summary tables and September 15, 2026 refreshed live CDE API have differing totals from this master file. The FBI README explains quality/methodology differences between master files, publication tables, and live dashboards. Therefore results here specifically identify the annual master file, rather than claiming to reproduce every FBI display.
Population denominators: Census NC-EST2025-SR11H, released June 2026, cells H49 (White-alone male), H92 (Black-alone female), H63 (non-Hispanic White-alone male). Race coding in administrative records and Census self-identification is not perfectly harmonized; race-alone denominators are used consistently.
Sources
- FBI 2025 annual release, August 14, 2026
- FBI downloads: Master File Downloads > Supplementary Homicide Report > 2025; help file; Crime in the United States Annual Reports > Expanded Homicide Tables > 2025.
- Census 2025 population table
- Census population release and definitions
The source archive, population spreadsheet, codebook, FBI methodology README, extracted results, and recalculation script are saved alongside this note.
Return to canonical video analysis. FBI methodology: Attachment (file not included in this release).
Recalculation: Attachment (file not included in this release).