Browse PlainCrime

How violent and property crime are reported in U.S. cities

City-level crime rates published on PlainCrime come from the FBI Uniform Crime Reporting (UCR) Program, which has compiled offense counts from local law enforcement agencies since 1930. Each year, thousands of police departments submit standardized counts of Part I offenses, murder and non-negligent manslaughter, rape, robbery, aggravated assault, burglary, larceny-theft, motor vehicle theft, and arson, to a central FBI database. The UCR program supplies the raw counts; population denominators come from the U.S. Census Bureau, and per-capita rates are calculated by dividing the offense count by population and multiplying by 100,000.

Violent crime, the focus of this ranking, comprises murder, rape, robbery, and aggravated assault. These four categories cover incidents involving force, threat of force, or completed physical harm against a person. Robbery requires an interaction between victim and offender even when no weapon is used. Aggravated assault, the most common violent crime nationally, captures attacks intended to cause serious bodily injury or involving a weapon. Murder counts include both completed killings and non-negligent manslaughter, but exclude negligent manslaughter, justifiable homicide, and suicide.

Property crime, burglary, larceny-theft, motor vehicle theft, and arson, appears far more frequently than violent crime in nearly every U.S. jurisdiction. The national property-crime rate typically runs five to seven times higher than the violent-crime rate. Burglary requires unlawful entry into a structure with intent to commit a theft or felony; larceny-theft covers simple taking of property without force; motor vehicle theft is specifically the theft of a vehicle; and arson is the willful burning of a structure or property.

Why per-capita rates matter more than raw counts

Comparing a city of 8 million residents to a city of 50,000 residents using raw offense counts produces misleading results: the larger city will always show more crime because there are simply more people to victimize. Per-capita rates (crimes per 100,000 residents) correct for population so cities of different sizes can be compared on the same scale. PlainCrime restricts its city rankings to jurisdictions with at least 25,000 residents to avoid extreme rate volatility, in a city of 5,000 people, a single robbery moves the per-capita robbery rate by 20 points, which is statistically meaningless.

Even with the 25,000-resident floor, per-capita rates contain meaningful noise. A city that hosts a major tourism destination, university, or commuter hub will record offenses against non-residents, while its population denominator counts only permanent residents. This inflates the rate. Conversely, a city where many residents commute elsewhere may record fewer victimizations than the resident count would predict, deflating the rate. We surface these caveats on each city's detail page so users can interpret figures in context.

Limitations of FBI UCR data

UCR participation is voluntary. While most large cities submit consistently, gaps appear, some cities report only certain categories, and some agencies fail to report in given years. The 2021 transition from the legacy Summary Reporting System (SRS) to the National Incident-Based Reporting System (NIBRS) reduced reporting coverage temporarily as agencies migrated systems. We flag known reporting gaps on each city page and use the most recent year of data available for each jurisdiction.

The UCR also captures only crimes reported to law enforcement. The National Crime Victimization Survey, run separately by the Bureau of Justice Statistics, consistently finds that property crimes are reported only 30–40% of the time and violent crimes 40–50% of the time. UCR rates should be read as reported crime, not total crime. They remain the best available comparable national dataset.