When America Orders an Autopsy (2024)
Only a small share of deaths are autopsied overall, but nearly every homicide is and almost no natural death is. A map of official suspicion.
Key findings
- Autopsies are rare overall but near-universal for suspicious deaths. Just 7.8% of all 2024 deaths were autopsied, but the rate depends almost entirely on how someone died.
- Homicide deaths were autopsied 98.3% of the time, versus only 3.5% of natural deaths.
- Suicides (49.8%) and accidents (44.4%) fall in between, autopsied roughly half the time.
- Autopsy also tracks age: 50.7% of deaths under 40 were autopsied, versus 0.7% of those 85 and older, because unexpected deaths concentrate among the young.
An autopsy is the system's way of asking "are we sure?" This measures how often one was performed on 2024 deaths, broken out by the manner of death and by age. The pattern is a map of suspicion: homicides are almost always examined, natural deaths almost never, and the young far more than the old. These are exact counts from the near-census of death records.
Autopsy rate, 2024
Autopsy rate by manner of death
| Share autopsied | Share autopsied |
|---|---|
| Homicide | 98.3% |
| Could not determine | 82.3% |
| Pending investigation | 73.9% |
| Suicide | 49.8% |
| Accident | 44.4% |
| Natural | 3.5% |
By manner and by age
| Group | Autopsy rate |
|---|---|
| Homicide (manner) | 98.3% |
| Could not determine (manner) | 82.3% |
| Pending investigation (manner) | 73.9% |
| Suicide (manner) | 49.8% |
| Accident (manner) | 44.4% |
| Natural (manner) | 3.5% |
| Under 40 (age) | 50.7% |
| 40 to 64 (age) | 18.3% |
| 65 to 74 (age) | 4.9% |
| 75 to 84 (age) | 1.9% |
| 85 and up (age) | 0.7% |
Autopsy rate is the share answered Yes among deaths with a Yes or No autopsy record. Exact counts; no sampling margins of error.
Data Sources & Methodology
The figures on this page are computed directly from the source data below. Narrative text accompanies the figures but never generates them.
CDC/NCHS Mortality Multiple Cause-of-Death Public Use File
- Vintage:
- 2024
- Access:
- BigQuery
- Cadence:
- annual
A near-census of U.S. death records for the year, drawn from death certificates and coded to ICD-10. Each row is one death, with the underlying cause and all contributing (multiple) causes, age, sex, race and Hispanic origin, month and day of death, place and manner of death (natural, accident, suicide, homicide), marital status, education, autopsy, and disposition. It supports analysis of what Americans die of and how that varies by age and demographics.
Every U.S. death is registered on a death certificate, which NCHS compiles, codes to the International Classification of Diseases (ICD-10), and releases as an annual public-use micro-file. Each record carries one underlying cause plus all contributing causes.
The file covers essentially all registered deaths in the data year, not a probability sample. Statistics are exact counts; there are no sampling weights and no sampling margins of error.
To protect confidentiality, the public-use file omits state and county of residence and occurrence. Analyses here are national and demographic (by cause, age, sex, race and Hispanic origin, month, manner, and place of death). State-level figures require CDC WONDER or the restricted-use file.
Causes follow ICD-10. Alongside the underlying cause, NCHS provides standard recodes (358, 113, and 39 groups) used in its publications, and the full list of entity-axis and record-axis contributing conditions.
- The public-use file has no state or county geography, so it cannot be mapped sub-nationally.
- Cause of death depends on how certifiers complete death certificates; coding conventions and ill-defined ("garbage") codes affect some categories.
- The final annual file lags roughly a year behind the data year; provisional counts (via CDC WONDER) are more timely but less detailed.
- Age is reported in several recodes; very fine demographic cross-tabulations can be thin for rare causes.
How these figures were produced
- Estimate type:
- Administrative near-census (exact counts)
- Weighting:
- None (not a weighted sample)
Suggested citation
Centers for Disease Control and Prevention, National Center for Health Statistics, 2024 Mortality Multiple Cause-of-Death Public Use File. Analysis by Matt George, itsmattgeorge.com. Retrieved 2026-10-24.