<p>The most established approach to legitimizing animal research worldwide is through Harm-Benefit Analyses (HBAs), which require that the expected benefits of research be compared to and outweigh the harms done to animals. For this process to be ethically sound, all relevant harms must be considered. However, HBAs leave one fundamental harm insufficiently considered: killing, i.e., the deliberate termination of animals’ lives as part of the research project. This article argues that the law’s ignorance of the harm of killing leads to its trivialization in practice and, consequently, undermines the ethical legitimacy of deadly animal research projects. While our argument applies broadly, we use Switzerland as an illustrative case example. We first disentangle, from an ethical standpoint, the harms that killings inflict on animals in research. Second, we show that most harms of killing, both to the inherent worth and wellbeing of animals, are not accounted for in law, allowing killings to be trivialized. Third, we propose that killing should always (a) be subject to ethical and legal scrutiny by animal research commissions, (b) have to pass the necessity test (i.e., killing should be permitted if, and only if, it is necessary to achieve the purpose of the experiment), and (c) be included in HBAs. Fourth, we consider common objections to our arguments, and, in concluding, show why recognizing the harm of killing is essential to upholding the ethical legitimacy of animal research.</p>