Autism Spectrum disorder (ASD) is a wide ranging developmental condition commonly recognised by impairments in social interaction, communication, repetitive behaviour and restricted interests. Deficits in imitation, sensory processing and motor control are also extensive, but conflicting findings mean a clear picture regarding the true nature of these is yet to be established. The aim of this thesis was to further investigate imitation with a specific focus on the effect of goals. It has been suggested that imitation occurs along two routes; a direct route for the imitation of meaningless actions i.e. goal-less, and a semantic route for meaningful or goal-directed actions. There is evidence to suggest that while individuals with ASD may have an impairment in goal-less imitation, goal-directed imitation is unaffected (Hamilton, Brindley & Frith, 2007). The experiments in the present work were based on the hypothesis that impaired goal-less imitation in ASD may be due to a problem with sensory motor integration in the direct visuomotor pathway for imitation (Tessari and Rumiati, 2004).The first experiment was conducted to ascertain whether movement kinematics differed between imitation during goal-directed and goal-less hand movements in a neurotypical control group. Participants observed and imitated hand movements of different speeds, while their movement was recorded with a motion sensor. Movement was modulated between the different speeds in the goal-less, but not goal-directed trials. These findings support the dual route model where visuomotor mapping occurs via the direct pathway during goal-less imitation but during goal-directed imitation there is greater reliance on representations of actions from long term memory.In experiment two, the same paradigm was employed, including additional movement types, and a group of adult individuals diagnosed with an ASD was compared to a control group. It was predicted that, unlike the control group, the ASD group would fail to modulate their movement in the goal-less condition, due to a disruption in the direct pathway. Eye movements were also recorded in this experiment, to ascertain if differences in gaze position or eye movements might influence the ability of the ASD group to imitate goal-less actions. The ASD group did not modulate their movement for goal-less imitation. In addition, the ASD group exhibited more goal-directed gaze, and less gaze toward the hand, than the control group. The ASD group also failed to maintain pursuit of the hand, which may have influenced the amount of information collected regarding the movement.The third study extended the investigation by introducing a training phase halfway through the experiment. Based on eye movement findings, it was hypothesised that drawing attention to the hand might increase the importance of the hand in goal-less trials, subsequently leading to increased tracking, and improved imitation. Perspective was also investigated to determine whether observing the action from an egocentric perspective would improve imitation. Movement modulation was not improved for either the post training or the egocentric condition. However, both training and egocentric viewing resulted in faster performance of movements while training also reduced the amount of movement variability; suggesting that these conditions made imitation easier. Findings strongly support the hypothesis that people with ASD are impaired in goal-less but not goal-directed imitation. These results are discussed in terms of bottom-up and top-down influences on imitation. The observed eye movement differences between the ASD and control group suggest eye movement abnormalities, and the finding that egocentric observation facilitates imitation indicates problems with visuospatial transformation during allocentric imitation. Finally, failure to pursue may be due to lack of attention to the hand movement, evidenced by no increase in pursuit after training. This thesis concludes with the suggestion that impaired goal-less imitation in ASD may result primarily from bottom-up low level visual processing and oculomotor control problems, combined with a high level attentional mechanism. Future studies need to address whether these are the primary causes of imitation difficulties, or whether there is a fundamental sensory motor integration deficit in ASD.