San Antonio Crime Map

The safest and most dangerous ZIP codes in San Antonio

The most-searched question about this data, and the one most sites get wrong. Rank by raw count and you just find the biggest ZIP code; rank by a naive per-capita rate and downtown crowns itself on the strength of the Riverwalk. Here it is done the fair way, by reported offenses per 1,000 residents over the trailing year, with the catches that ranking pointed out.

Two catches, and they are the whole point. The "safest" list excludes the military bases (JBSA Lackland, Fort Sam Houston): their reported crime is low only because SAPD is not their police force, a jurisdiction quirk, not safety. And some "most dangerous" ZIP codes are retail corridors: the rate divides real offenses by the few thousand who live there, while the crime is driven by the tens of thousands who shop there. A ZIP code is a mail boundary, not a neighborhood, and this is reported crime, not risk. Read it as a starting point, then look up the specific ZIP above.

San Antonio crime: the questions people ask

Is San Antonio safe?

It depends what you measure, and this map is built to let you. By the FBI's numbers San Antonio's violent-crime rate runs above the national average, and its property-crime rate is the highest of the five biggest Texas cities. But reported crime is heavily concentrated (half of it falls in about 16 of 62 ZIP codes) and heavily late-night, so a citywide label hides more than it tells. Look up your own ZIP rather than trust one number.

What is the safest ZIP code in San Antonio?

By reported offenses per resident over the trailing year, 78261 is the lowest, at about 0 per 1,000 residents. Military-base ZIP codes read even lower, but that is because SAPD does not police them, not because they are safer. A ZIP is a coarse mail boundary, so treat this as a starting point.

What is the most dangerous ZIP code or neighborhood in San Antonio?

By offenses per resident, 78202 ranks highest (about 187 per 1,000). By sheer number of offenses, 78207 has the most. Watch the trap: a couple of high-rate ZIP codes are retail corridors (the airport and the big malls), where the crime is driven by shoppers, not the few thousand residents the rate divides by.

What time of day does crime happen in San Antonio?

Violent-crime 911 calls (assaults, shootings, robberies, logged when they happen) climb through the evening and peak around 10 pm, heaviest on Sunday and the weekend. About 37% of them come in during the five hours from 9pm to 2am. Early morning is the quietest stretch.

What is the most common crime in San Antonio?

The most-reported NIBRS Group A offense is Simple Assault, followed by property damage and theft from motor vehicles. Most of what SAPD records is property and lower-level offenses, not violent crime.

What actually changed in 2026-08

Nothing did. Not one of the 51 ZIP codes we can test moved far enough from its own expected range to be distinguishable from ordinary month-to-month noise. Several of them changed by more than 20%, and that change still means nothing. Below is the evidence, including the ZIPs whose headline percentages look alarming.

Ranked by how far each ZIP code departed from its own expected count, not by percentage change. Expected is that ZIP's own 24-month baseline, adjusted for the seasonal pattern and for how many days the month has. The comparison is quasi-Poisson: offense counts cluster, so their variance genuinely exceeds their mean, and a plain Poisson test would call several ZIP codes "unusual" every single month just by misjudging the noise. A ZIP is only called unusual beyond 3.0 standard deviations, because testing 51 ZIP codes at once at the conventional 95% threshold would flag about 3 of them by chance alone. Withheld below 30 expected offenses.

Read the third and fifth columns together. That is the entire point of this panel: a ZIP code can move by half and still be doing nothing unusual, because a small denominator moves easily. Percentage change is the number every other crime map leads with, and it is the number most likely to mislead you.

Every offense runs on its own week

This is the single most misread thing in any crime map, and you can see it here. The City publishes the day an offense was reported, not the day it happened, and different offenses give that away in different ways. Each pair of charts is one offense type against an average day and an average month (the dashed line is 100). Sorted into what their weekly rhythm is really telling you.

We publish the weekday and month-of-year shapes because they are real and they are in the data. We are careful to tell you what they do and do not mean, because the obvious reading, that Monday is a dangerous day, is the wrong one. What the pattern actually measures is how and when each kind of crime reaches a police report. That is a genuinely useful thing to know, and almost nobody says it out loud.

When violent crime actually happens

The map's offense dates cannot tell you when a crime happened, only when it was reported. So this comes from a different place: 911 calls. A call about an assault, a shooting or a robbery is logged the moment it comes in, which is as close to the real time as this data gets. Each square is one hour of one weekday over three and a half years; the darker it is, the more violent-crime calls came in then.

So the honest version of "the most dangerous time" is late on a weekend night: violent-crime calls climb all evening, peak around 10 pm, and run heaviest Friday through Sunday. The quietest stretch is the early morning commute.

Read this as calls, not convictions. It counts 262,264 911 calls reporting assaults, shootings and robberies, by the hour they were made. A call is a report of a possible crime, which is a different thing from the confirmed offenses on the map above, and it comes from a different City dataset (Calls for Service). We use it only for timing, because it is the only place the timing honestly exists.

Arrests, for comparison: not the same thing as offenses

A second public dataset: 201,260 SAPD arrests across the same ZIP codes over the full 2023-01 to 2026-08 record. The table below shows the trailing 12 months (2025-09 to 2026-08), the same window as the map above. Do not divide arrests by offenses. They count different things. Arrests are mostly misdemeanors that are not in the Group A offense set at all, and an arrest can be made this month for a crime committed years ago, so an "arrests per offense" figure would be meaningless. We show arrests on their own, as their own measure.

ZIP codes ranked by arrests in the trailing 12 months. Felony share is the percentage of that ZIP's arrests that were felonies, not misdemeanors. A per-1,000 rate is withheld below 5,000 residents, the same rule the offenses table uses.

Arrest location is where the arrest was made, which we checked tracks the offense map closely, so it is honest to read this geographically. What it is not is a measure of how many crimes get solved. That number is not in the public data, and anyone who quotes you one from a source like this is guessing.

Take the data

Everything above is built from open data you can download here. The offenses CSV is one row per ZIP code, per month, per offense type, the shape you want for a spreadsheet or a notebook, and arrests come as their own CSV. No signup, no email, no attribution required to us. Prefer to pull it programmatically? There is a versioned public API with JSON, CSV, and an OpenAPI spec.

Download CSV JSON (indexed cube) JSON (ZIP geometry) Arrests CSV

The underlying records are published by the City of San Antonio under CC-BY (City of San Antonio). The City is the source and deserves the credit. If our processing of it is useful to you, a link back to this page is welcome but not required. Cite it as: BrandShyp, “San Antonio reported offenses by ZIP code”, derived from City of San Antonio open data (SAPD Offenses), built 2026-09-06.