Research focused on police officers’ decision-making in ambiguous use-of-force situations has yet to investigate the role that a suspect’s biological motion plays in unknown-object identification. The current study uses point-light displays to isolate the suspect’s motion and remove potentially biasing information (e.g., skin tone, facial expression, clothing). Experienced law enforcement officers and trainees (= 129) watched point-light display videos of an actor pulling either a weapon or a non-weapon from a concealed location in a threatening or non-threatening manner. After each video ended, participants indicated whether the object—which was not visible—was a weapon or a non-weapon. Results indicated that the speed and intent (e.g., threatening vs. non-threatening) with which the actor drew the object were significant predictors of officers’ responses. Officers’ law enforcement experience (i.e., years of service) was not a significant predictor of their response. This study has important implications for understanding why police sometimes make critical and costly errors in ambiguous use-of-force situations. We consider implications for police performance and developing improved training procedures.