Where Murders Go Unsolved

Reading the Map

Each county is shaded by a single number: the share of recorded homicides between 1976 and 2024 that were never cleared. Light tan means most murders are solved. Deep red means the majority go unsolved. The darkest concentrations sit in a band running from the Mid-Atlantic through the Midwest - Baltimore, Washington D.C., St. Louis, New Orleans, and their surrounding counties.

A homicide is "cleared" when law enforcement makes an arrest or closes the case by exceptional means - the death of a suspect, refusal of extradition, or similar circumstances. Clearance does not mean conviction.

What the Gray Counties Mean

Gray counties are not necessarily places without homicides. Of the roughly 3,100 U.S. counties, about 568 appear in the dataset with between 1 and 9 recorded homicides over the entire 1976-2024 period. Those counties are shown in gray because a rate calculated from fewer than 10 cases is statistically unreliable - a single case being cleared or not can swing the rate by 10 to 20 percentage points.

Gray can also reflect reporting gaps. The SHR is a voluntary program, and some local agencies have submitted data inconsistently or not at all. A gray county may have homicides that simply were not reported into the federal system - the absence of data is not the same as the absence of crime. The Murder Accountability Project has supplemented the raw FBI data with records from other sources, making this dataset more complete than a standard FBI download, but gaps remain.

Findings

Clearance rates have gotten worse over time

In the 1970s, about 25% of homicides went unsolved nationally. By the 1990s that figure had climbed to 32%, where it has roughly stayed through the 2010s. The shift tracks with the rise in stranger-on-stranger gun violence during the crack epidemic: homicides between people who knew each other - domestic violence, disputes between acquaintances - are far easier to solve than anonymous shootings on a street corner where no witness will talk. As the mix of homicides shifted toward the latter, the national clearance rate dropped and stayed down.

Firearms are harder to clear than other weapons

Across all years, gun homicides had a 31% unsolved rate compared to 25% for non-firearm homicides. This gap reflects both the anonymous nature of many shootings and the practical difficulty of forensic investigation - a stabbing or beating often leaves more physical evidence and implies closer contact between victim and suspect. Switch the map to "Firearms only" and the country gets noticeably darker, especially in dense urban counties.

The worst clearance rates are concentrated in cities

Among jurisdictions with at least 100 recorded homicides, the highest unsolved rates are almost entirely urban: Washington D.C. (60%), St. Louis city (54%), Boston's Suffolk County (54%), Baltimore (50%), and New Orleans (49%). Manhattan - which recorded over 48,000 homicides across the dataset - sits at 49% unsolved. These are not small-sample anomalies. They represent decades of accumulated cases that were never closed.

The 2020s appear better - but that may be an illusion

The 2020s filter shows a lower unsolved rate (~28%) than the 2010s. This is likely a reporting lag artifact rather than a genuine improvement. Recent homicides are still being investigated, and cases that will eventually be cleared have not yet been recorded as such in the dataset. Treat the 2020s numbers as a floor, not a ceiling.

Data and Methodology

The source data is the FBI's Supplementary Homicide Report (SHR), 1976-2024, as compiled and published by the Murder Accountability Project. The SHR is a voluntary program - law enforcement agencies submit case-level data on every homicide, including victim demographics, weapon, circumstance, and whether the case was cleared.

Records were aggregated to the county level using the FIPS code embedded in the SPSS version of the dataset (the CSV version replaces numeric FIPS codes with county name labels). The unsolved rate is:

Counties with fewer than 10 homicides in the selected period are shown in gray and excluded from the color scale to avoid unreliable small-sample rates.

Limitations

Clearance is not conviction. An arrest does not guarantee prosecution or a guilty verdict. The clearance rate measures whether law enforcement identified a suspect - not whether justice was served.

Voluntary reporting is uneven. Some agencies have submitted data inconsistently. County-level rates reflect only what was reported into the SHR system and may not capture every homicide that occurred.

County boundaries can mislead. Some jurisdictions - like St. Louis city and Baltimore city - are independent cities that are not part of any county. They appear as their own units in the data, which concentrates urban homicide patterns into a single visible shape on the map.