Recently, violent rhetoric and its use by political leaders in the United States has become a primary focus of elites, media, and researchers alike. This research unveils whether members of Congress use violent political rhetoric strategically in online communications, and if the rates at which they use it is consistent with intuitive and theoretical understanding. Using a dataset of Congressional e-newsletters from 2011 to 2020, I use a random forest classifier to predict the rates at which violent political rhetoric is used and analyze them over time and along party lines. I find that violent political rhetoric is prevalent among newsletters and has increased overtime in partial support of my theory, but that Democrats use it at greater rates than Republicans in contrast to what was postulated. These findings implicate our understanding of the role of violent rhetoric in politics as a driver of negative partisanship, polarization, and political violence.