This article examines people's interpretations of a recorded police-civilian altercation when they receive a priori information about a civilian's behavioral health and information about how an officer responded. In two experiments, MTurk participants (N = 771) were randomly assigned to receive a label for the civilian (schizophrenia, substance use, or no label) and information about the officer's response (use of force, de-escalation tactics, or no further information). Participants then watched body-worn camera footage of a real police encounter, answered questions about the officer's and civilian's behavior, and made punishment decisions. Participants who learned that the officer used force, compared to de-escalation techniques, interpreted the officer's behavior more negatively and the civilian's behavior more positively. Furthermore, participants who learned that the civilian used substances agreed the most strongly that the civilian should be punished, whereas participants who learned that the civilian had schizophrenia agreed the least strongly that the civilian should be punished. Participants attributed the civilian's behavior to personal shortcomings (e.g., poor life decisions) when they learned he used substances and to genetic factors when they learned he had schizophrenia. The belief that the civilian's behavior was because of personal shortcomings was associated with increased punishment. These findings have implications for the common belief that viewing police footage will ensure objective, fair, and just outcomes. We conclude with policy recommendations to improve analysis of police footage.