Regression in codebases is a significant problem that proportionally significant amounts of effort have already been spent addressing. Regression is a similarly large problem in the realm of security, yet de-facto standards and approaches for addressing the issue remain absent. Even when security programs have the proper staff, tooling, and budgets, they commonly struggle with ensuring that security holes remain fixed after they are initially patched. This talk will explore the application of a regression solution commonly employed in software development - unit testing - to fighting security regression. We will cover unit testing solutions that are both integrated into tested codebases as well as solutions that can test deployed codebases from a blackbox standpoint. Our talk will be aided by the release of an open-source software project built specifically to demonstrate how these practices can be employed in real-world scenarios and with re-usable core testing functionality that can be integrated into existing Python projects. Through this talk we hope the audience will leave with an understanding of the role that regression plays in security and how unit testing can be used as a tool to address security regression in both in-house codebases as well as untrusted third-party software.