What do we know about crime on New York City's subway system? A lot and a little.
A lot, in that crimes that happen there tend to be covered intensely in the media — especially pushings and homicides, including notorious incidents such as the arson homicide at the Coney Island-Stillwell Avenue station late last year. It's been widely reported that there were 10 homicides underground in 2024, which, while statistically rare for a system that serves nearly 2 billion passengers per year, is unusually high compared to other recent years. Similarly, violence on the subways, while also statistically rare by many measures — there's a little under one violent crime per million rides — has risen substantially since the COVID-19 pandemic as well as over the longer span of time, including on a per-rider basis.
But beyond data that compare the subway system to itself and that compare New York City's subways (generally unfavorably) to large transit systems in other global cities, there's little empirical research about crimes committed underground. As a result, researchers haven't learned much about its prevalence, where crimes concentrate, who the offenders are or what might be done to stop subway crime from happening.
Based on decades of research, we do know quite a bit about where, why and how crime happens, at least above ground. Those basic principles — which establish, for instance, the tendency of crime to concentrate in a small number of places, the small number of people responsible for the largest proportion of serious crimes and the effect of the physical environment on behavior — are important, because they point us toward the most effective approaches to preventing and deterring crime, such as hot-spot policing, focused deterrence and place-based interventions such as remediating blighted properties, cleaning up vacant lots and improving street lighting.
Is it the same underground, where quarters are close and people are especially mobile?
We analyzed public data on crimes, arrests and subway usage to understand trends in subway crime, what drives crime in the subway system and thus what the most effective approaches would be to stem it. Here are the most telling facts we discovered:
Crime trends
Subway violence, accounting for 4.2% of the total violent crimes in New York City, has increased since the pandemic but was increasing even before March 2020.
Despite declining ridership, subway violence increased markedly — by about 15% — from 2019 to 2024 and by 96% from 2014 to 2024. The early months of 2025 have seen a notable decrease in subway violence. (Throughout this report, the term "violent crimes" refers to the following offenses: murder, robbery, rape, and felony and misdemeanor assault.)
The increase in subway violence over the last decade — as well as violence in the city's public spaces on the surface — is driven by noninstrumental violence, meaning motivated by animus rather than by monetary motives. While robberies were nearly 1 in 3 violent subway crimes a decade ago, today they represent only 1 in 6 violent subway crimes. Felony assaults on the subway have more than tripled since 2009 even as robberies and grand larcenies are down by 18%. This tracks broader national trends as well as trends on New York City's surface, showing that violence is increasingly less likely to be due to instrumental or pecuniary motives and more likely to be due to personal animus. The rise in assaults could partly explain the recent uptick in murders which, while still rare, are up by an alarming amount.
The increase in felony assaults on the subway in 2024 is partially driven by increases in assaults on police officers. Assaults on police officers on the subway rose from 98 (17% of all felony assaults) in 2023 to 179 (31% of all felony assaults) in 2024. There has been a steady increase in the number of assaults on police officers since the pandemic.
While subway violence has increased, so has violence in public spaces on the surface. Subway violence is up by about 15% since 2019; street violence has risen by 16% over the same period. This suggests that whatever is happening on the subways reflects a broader phenomenon that has affected public spaces throughout the city.
Gun violence on the subway system remains very rare. Typically there are fewer than five shootings on the system per year among over 2 billion rides, which means that the risk of a shooting in the subway is likely to be substantially lower than it is on the street. This also suggests that large investments to keep guns out of the system may not ultimately yield correspondingly large improvements in safety.
Since 2019, violent recidivism — defined here as the share of individuals who were arrested for a given type of crime in two consecutive years — has risen among subway offenders, doubling from 3% to 6% over this period. While recidivism rose among all arrestees in New York City during this time period, in percentage terms, the increase is far larger for subway violence than among violent crimes in general.
Concentration of risk
While most crimes occur during peak hours of ridership, on a per-person or per-ride basis, the subways are safest during peak hours of use and are considerably more dangerous at night. While the risk of a rider being victimized is about 1 per million during rush hour, it is more than 20 times higher during the late night and early morning hours when fewer people are using the subway. This empirical regularity helps to explain the City's recent decision to deploy police officers on trains during the overnight hours.
Similar to the surface, crimes are heavily concentrated within the system. Of the city's 472 stations, the 30 stations with the largest number of violent incidents account for 50% of violent crimes. An important caveat here is that, in NYPD data, crimes are assigned to the nearest station even when they happen on trains, making it difficult to understand exactly where the crimes are happening — in the stations themselves, on mezzanines or on trains pulling into a given station. Many of these stations are the major transit hubs which serve the largest number of passengers. However, the stations where the risk per passenger is greatest are quite different. These stations tend to be used less frequently than the major hubs and to be located in the outer boroughs. The amount of violence at these stations, year to year, is not easy to predict.
The most vulnerable New Yorkers, the elderly in general and elderly women in particular, represent a very small share of the victims — between 1% and 3%, depending on how the vulnerable group is defined. Two-thirds of the victims of subway violence are male, and people between the ages of 25 and 44 are most likely to be the victims of subway crime.
Subway offenders
Subway crimes, like other crimes in New York City, are being committed by individuals who are older than many people might suspect. Just 36% of arrestees for subway violence are under the age of 25, and only 14% are juveniles under the age of 18. Half of arrestees are between the ages of 25 and 44, and 13% are over the age of 45.
Violent subway offenders are getting older, not younger, which suggests that recent Raise the Age legislation is unlikely to be a major driver of the recent increase in violence. Today, individuals arrested for violent crimes in public spaces are, on average, eight years older than they were 18 years ago. Violent subway arrestees are still slightly younger than violent street arrestees, but the difference in ages has diminished rapidly.
Individuals with a violent subway arrest have more extensive criminal histories than those arrested for a violent street crime. Violent subway arrestees are, for example, 6 times more likely to have other arrests for a violent subway crime than other arrestees. Likewise, while fare beating arrests are, in general, not very predictive of future subway violence, when compared to other arrestees, there is a greater propensity for violent subway arrestees to have a prior arrest for fare beating — potentially indicating that this is a population that spends more time in the subway system.
Individuals with the largest number of arrests for subway crimes and subway violence are more likely than other arrestees to be homeless and mentally ill. One NYPD analysis of high-volume subway offenders found that 41% were homeless at the time of their most recent arrest. Nearly two-thirds of those with the largest number of subway arrests in 2022 and 2023 had a documented history of homelessness or mental illness. Among those who also had an arrest for subway violence, 89% had a documented history of homelessness, mental illness or both.
Disorderly conduct, fare beating and subway violence
Since 2019, citizen calls about disorderly conduct and arrests for disorderly conduct have more than doubled on the subway. However, less than 1% of 911 calls reporting disorderly conduct on trains result in an arrest. In 2024, for example, only 221 arrests were made on the subway for disorderly conduct, aggravated harassment, criminal mischief, criminal nuisance and unnecessary noise combined, out of 28,329 calls concerning disorderly behavior among individuals or groups.
Although many violent subway offenders do have a history of skipping fares, the vast majority of people caught fare beating are not actually arrested but instead receive a summons. Of the small group who do get arrested for fare evasion—around 7% of those caught—statistics show they are less likely to commit violent offenses later on than other arrestees, and overall pose a lower criminal risk. For example, fewer than 1 in 100 of those arrested for fare beating from 2016 to 2021 were subsequently arrested for a violent subway crime over the next two years. While this is a higher risk than violent crime arrestees on the surface, it is a far lower risk than among individuals who are arrested for disorderly conduct on the subways. Interestingly, a drug or disorderly conduct arrest below ground is more predictive of future subway violence than an arrest for a violent crime committed above ground.