In most major cities in America, shooters who kill someone with a gun have more than a 1 out of 2 chance that their crime will go unpunished. If the victim survives, the shooter’s chance of getting away with it increases to 2 out of 3.

Every year, hundreds — even thousands — of nonfatal shootings never even get assigned to a detective. The shooters are left free to shoot again.

In January, The Trace and BuzzFeed News published an investigation detailing these findings: “Shoot Someone In a Major U.S. City, and Odds Are You’ll Get Away With It,” and “5 Things to Know About Cities’ Failure to Arrest Shooters.”

To do this reporting, we sent public records requests to dozens of the nation’s largest police and sheriff’s departments. We requested detailed data on homicides and nonfatal shootings. We asked for information such as victim and suspect names, demographic details, weapon, case status, location, and motive. Because many departments don’t classify “nonfatal shooting” as a specific crime category, we expanded our requests to include all rapes, robberies, and aggravated assaults.

Some agencies gave us everything we asked for and more. Others dug in their heels on basic details, like weapon type and location. Case status was, by far, the toughest information to obtain, particularly for non-fatal shootings, which don’t get as much scrutiny from politicians and the media. Negotiations, appeals, and in some cases, litigation stretched on for months. Several agencies still haven’t given us any data.

Because the broader data may be useful to reporters, researchers, policymakers, advocates, and anyone else interested in examining America’s gun violence epidemic, we are posting it here. (We posted the standardized data, methodology, and code used for our stories at the time of publication.)

We also compiled tips for making public records requests for data, including suggested language and responses to common agency objections. (It’s always a good idea to check your city’s open data portal before filing a request, or the Police Data Initiative if you’re compiling data from multiple agencies.)

We’d like to know if you use this data. Please send your name, title, and contact information to [email protected] if you download the data. If you do any work that builds off this material, please share it with us. With your permission, we may highlight your work.

— Sarah Ryley

Download raw data from 56 police and sheriff’s departments

The following links will download zip-files of unmodified data or PDF files, and in some cases, documentation like code definitions and communication with the agency. We’re sharing incident-level data on murders from 56 agencies, nonfatal shootings from 20 agencies, aggravated assaults from 21 agencies, robberies from 12 agencies, and rape from 6 agencies. The timespan covered by each agency varies, but is generally from 2005 through late 2017.

We caution that the unmodified version of the data is not always the most accurate or usable version. For instance, we often found big differences between the number of incidents in the datasets and the number of incidents that the agency reported to the public. Agencies also vary in how they classify incidents and case status.

The 22 datasets that we standardized for our published stories underwent extensive research and fact-checking. We detail this process, the potential pitfalls, and the modifications we made to the data in our methodology. We encourage you to read it before using any of the information posted here. Any additional documentation that we’ve shared is intended to assist you in understanding the data. All questions for clarification and fact-checking should be directed to the relevant agency. The Trace and BuzzFeed News make no representations or warranties as to any third party use of this information.

The data was obtained by Trace journalists Sarah Ryley and Ann Givens; Trace data editor Daniel Nass; Trace fellows Sean Campbell and Francesca Mirabile; Trace contributors Anna Boiko-Weyrauch and Nolan Hicks; and BuzzFeed News data editor Jeremy Singer-Vine. We could not have obtained the data from the Newark Police Division, New York Police Department, and numerous California police and sheriff’s departments without pro bono representation from Jean-Paul Jassy and Kevin Vick of Jassy Vick Carolan LLP, Gideon Oliver of Gideon Law, Adam Marshall and Katie Townsend of the Reporters Committee for Freedom of the Press, and C.J. Griffin of Pashman Stein Walder Hayden.

Download standardized data from 22 police departments and the FBI

Police Department Data: We used internal data from 22 police departments to examine present-day trends in arrest rates for murders and other violent gun crimes. The standardized data, methodology, and code are available here.

Baltimore Police Department Shooting Data: We used victim and suspect data on nearly 3,500 fatal and non-fatal shootings that took place in Baltimore to analyze how shootings are linked through shared victims and suspects. The standardized data, methodology, and code are available here.

FBI Data: We used three major FBI datasets — the Supplementary Homicide Report, the National Incident-Based Reporting System, and Return A — to examine historical trends in unsolved shootings. The standardized data, methodology, and code are available here.

Read stories based on this data

Shoot Someone In a Major U.S. City, and Odds Are You’ll Get Away With It

5 Things to Know About Cities’ Failure to Arrest Shooters

In New York, the Neighborhood You’re Shot in May Determine Whether You Survive