INTRODUCTION: Law enforcement officers typically use lethal force judiciously, however mistake-of-fact shootings can occur. An officer's ability to identify the suspect's intent and what they are holding can be compromised by degraded visual conditions (e.g., low light) and the rapidity of movement. Therefore, the ability to infer intent from a suspect's underlying biological motion is likely to affect decision making. This prompts the following questions: (1) Can officers distinguish the motion of someone drawing a weapon from that of a person presenting a non-weapon? (2) Can experienced officers distinguish between these motions more accurately than novice officers? PURPOSE: To investigate whether draw type and draw speed affect officers' ability to distinguish a weapon from a non-weapon using biological motion as a cue. METHOD: An optical motion-capture system will be used to create dynamic point-light displays of a suspect actor drawing either a weapon or a non-weapon with either the intent to "attack" or to "comply." These stimuli will be used to examine the effect of draw type (Study 1) and draw speed (Study 2) on officers' ability to determine if the suspect is pulling a weapon or a non-weapon. Law enforcement officers and recruits will view the point-light display stimuli and determine whether the suspect actor is drawing a weapon or a non-weapon. The discrimination data from Study 1 will be used to calculate hit and false alarm rates for each draw type, which will be used to calculate independent measures of sensitivity (α) and response bias (β). These will each be analyzed using a 2 (experience level) x 2 (draw type) mixed ANOVA. The discrimination data from Study 2 will be analyzed using a 2 (experience level) x 2 (draw type) x 2 (draw speed) mixed ANOVA. RESULTS: Experienced officers are expected to be more sensitive than recruits in their ability to discriminate a weapon from a non-weapon and recruits are expected to have a more liberal response bias (i.e., tendency to respond "weapon") across all conditions. An interaction is predicted such that recruits are less sensitive for comply- versus attack-type draws. CONCLUSION: Law enforcement use-of-force training is often conducted in an ad hoc manner, therefore empirical knowledge -- such as that collected from this study -- regarding factors that impact use-of-force decisions may be used to develop validated training that could be offered nationally, helping to reduce the number of mistake-of-fact shootings involving unarmed suspects.