Under conditions of physiological stress, officers are sometimes required to make split-second life-or-death decisions, where deficits in performance can have tragic outcomes, including serious injury or death and strained police–community relations. The current study assessed the performance of 122 active-duty police officers during a realistic lethal force scenario to examine whether performance was affected by the officer’s level of operational skills training, years of police service, and stress reactivity. Results demonstrated that the scenario produced elevated heart rates (i.e., 150 beats per minute), as well as perceptual and cognitive distortions, such as tunnel vision, commensurate with those observed in naturalistic use of force encounters. The average performance rating from the scenario was 59%, with 27% of participants making at least one lethal force error. Elevated stress reactivity was a predictor of poorer performance and increased lethal force errors. Level of training and years of police service had differential and complex effects on both performance and lethal force errors. Our results illustrate the need to critically reflect on police training practices and continue to make evidence-based improvements to training. The findings also highlight that while training may significantly improve outcomes, flawless performance is likely not probable, given the limits of human performance under stress. Implications for the objective reasonableness standard, which is used to assess the appropriateness of force in courts of law, are discussed.