crime.pulse
Last updated
View

2016–2025 · 5 published indicators

New York
crime trends.

NYPD complaint data for assault, auto theft, burglary, robbery, and grand larceny.

Sources & methodology
5 published indicators▼ 4.2% vs 2024
176,384incidents · 2025
Peak
184,068 · 2024
Low
137,970 · 2020
vs 2016
+15.5%

By category · 2025 count

Trends across 5 indicators

Select an indicator to explore it on the map and charts. Each mini-chart has its own scale; compare the labelled values.

Citywide · 78 precincts

Assault count, 2025 · by precinct

Local boundaries shaded by one category at a time, in the active unit, for the selected year. Hover for a quick read; click to open the full precinct view.

Count by year

counts · 2016–2025 · baseline = all categories
052.9K105.8K158.8K211.7K'16'17'18'19'20'21'22'23'24'252020 · pandemicAll categories
● AssaultAll categories (baseline)
View chart data

Full years only; incomplete YTD totals are excluded. Not available means a value is missing, not zero.

New York-wide yearly trend · incident counts · 2016–2025
YearAssaultAll categories (baseline)
201673,260152,774
201771,631146,661
201873,470147,186
201973,480146,378
202064,265137,970
202171,501151,021
202281,207179,658
202386,240182,998
202492,497184,068
202589,799176,384

How the series is built

Each year uses one NYPD complaint file: the historic file when it has at least as many complaint-from-date rows as the year-to-date file, otherwise the year-to-date file. The two files are not added together. 2026 comes from the year-to-date file and is excluded from the ten-year comparison.

Grand larceny is the felony theft cut, value over $1,000. Petit larceny stays out. Assault is felony assault plus assault in the third degree. Harassment, sex offences, and homicide stay out.

Population

Citywide rates use U.S. Census Bureau city estimates. 2016–2019 are from the 2010-based vintage and 2020–2025 from the 2020-based vintage, so the denominator changes base in 2020. Precinct rates use 2020 Census block-group population assigned by each block group’s internal point. Precinct 22 is Central Park; its residential population is tiny, so its rate is not comparable with residential precincts.

0.0% of these indicator reports could not be placed in a local area and stay in the city totals only.

Reading the data

Three numbers that need context

!
Auto theft fell 6% in 2025 after peaking in 2023. This is the change in recorded grand larceny of a motor vehicle, not a prediction of future risk.
×
Harassment, sex offences, and homicide are excluded. Grand larceny is New York’s over-$1,000 cut, not a $5,000 theft-over cut.
i
Counts by default, rates one toggle away. Citywide rates use annual U.S. Census Bureau city estimates. Precinct rates use a fixed 2020 Census allocation. Precinct 22 covers Central Park, so a small residential denominator makes its rate much higher than a residential precinct with the same count.

Data summary for New York-wide

New York-wide: rate per 10,000 residents by category, 2016–2025. A dash means unavailable.
YearAssaultAuto TheftBurglaryRobberyGrand Larceny
201686.57.515.418.352.7
201784.96.714.416.651.2
201887.56.414.015.451.9
201988.16.513.116.151.8
202073.410.417.915.140.9
202184.612.315.316.450.1
202297.116.518.920.961.5
2023102.318.716.320.159.6
2024107.616.515.219.355.6
2025104.615.614.817.553.0

About this data

How to read these figures

What does “crime” mean for New York?
Crime by precinct means the indicators New York Police Department publishes for New York (2016–2025). They are reported incident or offence counts, not arrests, charges, or a safety score. The same heading in another city is not the same offence.
Where do the numbers come from?
Police figures come from New York Police Department. Area boundaries come from NYC Department of City Planning · police precincts. Rates, when a denominator exists, use U.S. Census Bureau · population and are shown per 10,000 residents. A missing year, indicator, or boundary stays missing.
How often is this updated?
The site is updated periodically, with no fixed public schedule and no live feed. Comparable charts use full years. A year-to-date control, when the page has one, labels the partial year separately.
What does this not show?
Drug offences and overdose deaths are separate series, not part of these precincts. There is no single cross-city crime index. Counting rules and source links are on the New York about page.

Full methodology →