Over the past two decades, autism research literature has grown dramatically, yet the attention placed on individuals with the greatest degree of impairment has not kept pace. One reason for this discrepancy is the lack of effective and appropriate assessment tools for this population. The challenge facing scientists developing assessments is the limited behavioral repertoire of nonverbal individuals with autism. Jeffrey Munson and his colleagues at the University of Washington are making use of a new technology that measures full-body movement and eye gaze within a game-like setting in order to assess motor planning and imitation ability in individuals with severe autism. These tasks are accessible to nearly all individuals, from young children to adults, regardless of their degree of intellectual or communication impairment. The setup involves a Microsoft Kinect motion-sensing device, which translates the individual’s body movements into movements made by the on-screen avatar. Munson and his team have developed two motor-planning tasks that vary in the required speed of movement (popping balloons for quick movements versus balancing blocks for slow movements) and two imitation tasks that vary in social context (imitating a human-like avatar versus following an image of an airplane). They plan to present these tasks to adolescents with severe autism on multiple occasions, in order to assess the learning process as it unfolds. Repeated participation in these tasks may lead to improved performance in motor planning efficiency, accuracy and imitation ability over time. The researchers plan to explore how performance on these tasks relates to autism symptoms, intellectual and adaptive functioning, and receptive language, using traditional measures.