Philosophers of self-defense debate whether it can be morally permissible to kill an aggressor who only appears to threaten you. In developing moral theories of self-defense, these philosophers sometimes make (untested) conjectures about what most people believe about self-defense. This paper aims to explore lay judgments on this issue. To do so, we conduct three pre-registered experiments manipulating the actuality of a threat. Across abstract and concrete scenarios as well as within-subjects and between-subjects designs, results consistently show that laypeople judge certain self-defensive killings morally permissible regardless of whether the aggressor poses a genuine threat or a merely apparent threat. These findings oppose Objectivist views on self-defense, which hold that self-defense is only permissible when facing a genuine threat. Instead, they support Subjectivism and what we call the “It’s Complicated View”, both of which hold that apparent threats can justify lethal self-defense (albeit with possible variation in permissibility ratings).