How accurate are crime maps? They are only as complete as the police reports behind them. Crimes that are never reported are absent, and areas where an agency has not supplied data can appear blank.
Those two gaps, plus some display quirks, explain most of the surprises people meet on a map. For the basics of filters and searching, see how to use a crime map. This guide covers what the pins cannot tell you.
What a crime map actually counts
A crime map plots incidents that someone reported to law enforcement. The Council on Criminal Justice notes that the FBI’s Uniform Crime Reporting program tracks crimes reported to law enforcement authorities, while the national victimization survey also captures offenses that never reached police. A pin is therefore a record of a report, not a measurement of everything that happened on that street.
That distinction decides how to read a map. It is a reasonable guide to what police were told and recorded in an area. It is a weak guide to the full amount of crime, especially for offenses that victims rarely report. The path from a report to a pin, including which agencies feed which maps, is covered in where does crime map data come from.
Unreported crime and what maps leave out
Crime maps do not include unreported crimes, because there is no police record to plot. Vera Institute of Justice states that many experiences of crime are never reported to the police and, therefore, are not counted in FBI statistics. The Council on Criminal Justice adds that most crimes in the U.S. are never reported, and that less severe crimes are reported at even lower levels.
Serious offenses are not fully captured either. Vera reports that nearly half of serious violent crimes go unreported to the police. The Bureau of Justice Statistics runs the National Crime Victimization Survey, which collects information on victimizations both reported and not reported to police. That survey is the main national check on what police-based data miss.
In practice, a quiet block on a map may mean little happened. It may equally mean that residents rarely report incidents there. The map cannot distinguish the two cases, so treat the pin count as a floor rather than a total.
Jurisdiction boundaries and missing areas
A nearby area is usually missing data because its agency has not supplied complete records to the map or to the national system behind it. Maps draw on agency submissions, so coverage follows agency reporting rather than the street grid. Vera reports that in 2021 only 37% of 16,903 agencies reported complete data to the FBI.
National coverage figures look better than local ones. USAFacts reports that agencies that reported data in 2024 represent about 95.3% of the population of U.S. metros, and that FBI data include 337 metros with at least 80% of the metro population covered. The same analysis covers 217 cities with at least 100,000 people in 2024 and notes that not all cities have reported data. A strong national share can still leave a particular suburb, town or district thin.
This matters most when you compare two neighbouring places. If one side of a boundary line shows many incidents and the other shows few, check whether both agencies supply data to the same map before concluding that one area is safer. The adjacent-area gap is a reporting difference first and a crime difference only afterwards.
Display limits and map messages
Even when data exist, the map interface can show only part of them. The public map at CrimeMapping.com displays the message “Max record count of 500 reached. Zoom in or adjust filter.” When you see it, the pins on screen are a subset of the incidents in view, so zoom in or narrow the filter before reading the density.
The same interface shows “No data provided for this map extent” and “There are missing records.” These notices are the clearest signs of a coverage gap. An empty map under that message means no data were supplied for that view, which is different from a period without incidents.
Location is the other display issue. A pin sits on a street map, and agencies publish their own notices about how precisely locations are shown. Read the disclaimer on the specific map you use before treating a pin as an exact address, and treat clusters at intersections as approximate.
Limits of the survey that measures unreported crime
The victimization survey is the best national yardstick for unreported crime, but it has its own limits. It is a household survey, so its estimates carry sampling error and do not map down to a single neighbourhood. The table summarises the limits that matter when you use it alongside a crime map.
| Limit | What it means | Planning use |
|---|---|---|
| Self-reported | Respondents describe their own experiences, collected annually from January 1 to December 31. | Expect recall and reporting differences from police records. |
| Recall period | Estimates rest on crimes respondents experienced during the prior 6 months. | Do not read it as a day-by-day record. |
| Survey year | Crimes are classified by the year of the survey, not the year of the crime. The 2024 survey covers July 1, 2023 to November 30, 2024. | Do not line survey years up exactly with map dates. |
| Response rate | 60% of eligible households completed an interview in 2024. | Allow for households that did not respond. |
| Who is excluded | Persons living on military bases, in institutional settings, and persons who are homeless. | Some residents are outside the estimate. |
| Sampling error | Estimates that appear different may not be statistically different. Larger standard errors mean less precision. | Treat small gaps between estimates as inconclusive. |
Figures from the Bureau of Justice Statistics National Crime Victimization Survey page.
None of this makes the survey unreliable. It means the gap between police-reported and total crime is estimated with a margin of error, and the estimate should shape your caution rather than give you a precise correction factor.
Reading a map with these limits in mind
Use the map for patterns in reported incidents and cross-check anything you plan to act on. A short routine keeps the limits in view.
- Read the notice first. Check the map’s disclaimer and any banners about missing records or a record cap before looking at pins.
- Confirm the agency covers the area. Find out which department supplies the data for each place you compare, especially across a boundary line.
- Zoom in and filter. A capped view can hide incidents, so narrow the area or category until the cap message disappears.
- Compare like with like. Use the same date range and offense types for every area, and expect low reporting for less severe offenses.
- Try more than one source. Different maps draw on different agencies, and our comparison of the best crime map websites shows what each offers.
Frequently asked questions
Can I compare 2021 crime data with other years?
Not safely. USAFacts states that, due to methodology disruptions, 2021 FBI data are not comparable to other years, so a jump or drop around that year may reflect reporting changes.
Do police statistics and the survey disagree because one is wrong?
Usually not. The Council on Criminal Justice says much of the difference between the two sources stems from variations in methodology rather than inaccuracies. It also notes that victimization rates sit within a margin of error.
Does the victimization survey include businesses?
No. The Council on Criminal Justice notes that some victims, such as businesses, are accounted for in police-based UCR data but are not included in the survey.
How large is the survey behind unreported-crime estimates?
The Bureau of Justice Statistics describes a nationally representative sample of about 240,000 persons in about 150,000 households, surveyed each year.
