Headlines about subway crime in New York tend to run in two directions: either “the subway is getting dangerous again” or “subway crime is at historic lows.” Both have been true in 2026, sometimes in the same week. What most of those headlines don’t give you is the practical breakdown: which parts of the system carry the most risk, and during what hours. That’s what this article is for.
Here’s what NYPD data, directly from Chief of Transit Joseph Gulotta, shows as of mid-April 2026: subway crime is down 5.7% year-to-date, with more than 1,700 fewer reported crimes compared to the same period in 2025. That’s the headline number. But it doesn’t tell the full story of when and where that risk actually concentrates — and for most New Yorkers, the “when” matters a great deal more than the overall total.
The 2026 Picture: A Rocky Start, Then a Correction
January 2026 was the outlier that drove a lot of the alarming coverage. According to Police Commissioner Jessica Tisch, transit crime in January jumped 15% above the same month in 2025 — a meaningful spike after 2025 had closed as the safest year in the subway system since 2009. Department data showed major transit crimes running 17% higher through the first two months of the year, with robberies up 58% (60 incidents versus 38 in the same period of 2025) and felony assaults up 9%.
Why did January go sideways? NYPD officials pointed directly to the weather. The extreme cold snap pushed a higher-than-usual number of people — including individuals experiencing homelessness and mental health crises — into the underground system for extended periods. Chief of Transit Gulotta noted the surge was disproportionately driven by youth-on-youth robberies, a pattern that spiked in the confined, predictable environment of subway platforms and cars during the coldest weeks.
The NYPD response was rapid. At the end of February, the department surged 150 additional officers into the transit system. By March, crime fell across every transit category. Commissioner Tisch announced at the end of Q1 that transit crime was down 1.3% for the quarter overall, with felony assaults — historically the most difficult category to reduce — down 6.6%. By mid-April, with an additional cohort of officers deployed, the year-to-date figure had improved to down 5.7%.
Context matters here: 2025 was itself one of the best years on record, with overall major transit crime down 5.2% from 2024 and 14.4% below pre-pandemic 2019 levels. So being down 5.7% from that baseline is a real achievement, not a statistical rebound to easy numbers.
What the Data Actually Says About Risk: It’s All About the Hours
The question most riders actually want answered isn’t “how many crimes happened this year?” It’s “am I safe on this train, right now?” NYPD data provides a surprisingly precise answer, and it may change how you think about subway safety.
During rush hour, the risk to any individual rider is approximately 1 violent crime per million rides. That’s an extraordinarily low number. The New York City subway carries about 3.5 million riders on a typical weekday. At that rate, you would need to commute for roughly the next 2,700 years before statistically encountering a violent crime. Rush hour is genuinely, measurably safe.
Late night and early morning hours are a different story. According to research drawing on NYPD CompStat data, the per-rider risk of violent crime during overnight hours — roughly 1 a.m. to 5 a.m. — is 20 to 25 times higher than during peak commute hours. That’s not a small difference. For practical purposes, a 1 a.m. subway ride carries a fundamentally different risk profile than a 8 a.m. one, even on the same line and platform.
The NYPD understands this. Since Governor Hochul’s 2024 initiative, the department has deployed uniformed officers on-board every subway train in service between 9 p.m. and 5 a.m. The $77 million commitment for enhanced overnight subway patrols in 2026 continues and expands that coverage. This is a direct attempt to close the late-night gap that the data has identified.
Where Crime Concentrates: 30 Stations, Not 472
Another piece of the picture that aggregate crime statistics obscure: crime in the subway system is not spread evenly across its 472 stations. Half of all violent crimes in the transit system occur at just 30 stations — about 6% of total stations. By raw count, those 30 are generally the system’s busiest hubs: Times Square-42nd Street, Atlantic Avenue-Barclays Center, Jamaica Center, and similar high-throughput locations.
But here’s the important nuance. Those 30 stations are not where any individual rider’s risk is highest — they’re just where the absolute number of incidents is highest because so many people pass through them. On a per-rider basis, the math looks very different. The highest risk to a given rider, by far, is at smaller, outlying stations that have lower ridership — particularly during off-peak hours when fewer people are around and response times are slower. An incident at a quiet end-of-line station at 2 a.m. is far more likely to result in a completed crime than one attempted at a crowded midtown hub at noon.
This distinction matters for how you think about safety decisions. Avoiding a busy transfer station because it has a high crime count in absolute terms may not actually reduce your risk — especially if the alternative route takes you through smaller stations at late hours.
The Grand Central Incident: What April 11 Revealed
On April 11, 2026, a man named Anthony Griffin attacked subway riders at Grand Central Station starting at 9:40 a.m. He began on the 7 line platform, slashing an 84-year-old man in the head. He then moved to the 4/5/6 platform, where he cut a 65-year-old man (who suffered a skull fracture) and slashed a 70-year-old woman’s shoulder. Officers confronted and shot Griffin, ending the attack.
NYPD Chief of Transit Gulotta cited this incident directly when explaining how the department is evolving its officer placement strategy. Grand Central is one of the system’s most heavily patrolled locations — the attack illustrates that no deployment strategy eliminates risk from determined individuals experiencing acute crises. What the presence of officers did do was limit the duration of the attack and ensure an immediate armed response.
Three murders have occurred in the transit system so far in 2026, compared to zero at the same point in 2025. Each has involved circumstances — a Bronx platform shooting, a track-push on the Upper East Side — that reflect the hardest category of transit crime to deter through standard patrol presence alone.
Your Practical Safety Map for 2026
Translating all of this data into daily decisions:
Rush hour commutes (roughly 7–10 a.m. and 4–7 p.m.) are the statistically safest time to ride the subway. Per-rider crime risk is at its lowest. Crowded platforms and cars mean more witnesses and faster response. If you’re choosing when to take an optional trip by train, daytime and peak hours are the safest window.
Late-night travel carries measurably higher risk per rider. The NYPD’s overnight officer presence on every train between 9 p.m. and 5 a.m. has helped — but the risk differential compared to daytime is real and documented. If you’re taking the subway after midnight, staying near other riders, choosing well-lit car positions (near the conductor’s car), and being aware of your surroundings on the platform matters more than it does during peak hours.
Busy hub stations, while higher in absolute crime counts, are not where per-rider risk is greatest. The counterintuitive finding is that smaller, outlying stations at off-peak hours carry the highest individual risk. If you’re regularly commuting through a quieter station at off-hours, that’s the scenario that matches the highest-risk profile in the data.
The early-2026 robbery spike was real but concentrated. Youth-on-youth robberies drove much of the January surge, disproportionately affecting younger riders (particularly teens and young adults). If you’re in that demographic, situational awareness around phone and headphone use on platforms — particularly at night — directly addresses the most common robbery vector identified in the CompStat data.
Grand larceny (theft) remains the highest-volume category. Pickpocketing and phone snatching occur at far higher rates than violent crime, concentrated in crowded situations — busy platforms, packed rush-hour cars, and distracted riders. This is the category most likely to affect any given rider over the course of a year.
The 2025 Baseline to Keep in Mind
When you hear that subway crime is up 15% year-over-year (which January 2026 briefly was), or down 5.7% (which April 2026 is), the comparison baseline matters. 2025 closed as the safest year in the NYC subway system since 2009 — and 2009 itself was the modern low point before the pandemic disrupted everything.
Put differently: the New York City subway today operates at crime levels that were considered the floor of what was possible back in the early 2010s. The per-rider rate of major crimes has fallen roughly 30% since 2021 even as ridership recovered. The trend line over a five-year view points strongly in the right direction.
That doesn’t mean every ride is risk-free or that incidents like the Grand Central attack don’t happen. It means that for most riders, on most routes, at most hours — the statistical reality of taking the subway is considerably safer than the news cycle suggests. The risks that remain are real, but they cluster in specific conditions: late hours, low-ridership stations, and targeted theft in crowded moments of distraction.
Knowing that changes what “be careful on the subway” actually means in practice — and that’s a more useful instruction than a generic warning based on this week’s headlines.
Sources: NYPD Chief of Transit Joseph Gulotta via amNewYork, April 15, 2026 (amny.com); NYC Police Commissioner Jessica Tisch Q1 2026 briefing via NY1, April 2, 2026 (ny1.com); Governor Hochul press release on 2025 transit safety, December 2025 (governor.ny.gov); NYPD Transit-Bus Crime Reports (nyc.gov/nypd); NYPD CompStat 2.0 (compstat.nypdonline.org).
