In "The Immorality of Punishment", Michael Zimmerman attempts to show that punishment is morally unjustified and therefore wrong. In this response, I focus on two main questions. First, I examine whether Zimmerman's empirical claims-concerning our inability to identify wrongdoers who satisfy conditions on blameworthiness and who might be reformed through punishment, and the comparative efficacy of punitive and non-punitive responses to crime-stand up to scrutiny. Second, I argue that his crucial argument from luck depends on claims about counterfactuals that ought to be rejected. I conclude that though his arguments are powerful, they fall short of his ambitious aim of demonstrating that punishment is always seriously wrong.