Firearms are responsible for the majority of suicide deaths in the U.S. In an effort to identify handgun purchasers at risk of suicide, researchers retroactively tested the reliability of a machine learning algorithm on a database of handgun purchases in the state of California from 1996-2015. Purchasing a revolver-style handgun, female gender, white race/ethnicity, and no previous firearm transaction were associated with firearm suicide within one year of purchase. Overall, the most important variables for accurately predicting firearm suicide within one year were handgun type, race/ethnicity, purchaser age, and month of purchase. These findings shed light on alternative methods of suicide prevention and may have implications for advocates and policymakers. To learn more, see the study in JAMA.