Autism is a developmental disorder in which, among other things, specific deficits are found at several psychological tasks. Autistic children show lower performances at task that demand the processing of wholes. In contrast, they perform rather well at tasks that demand the processing of details. These results have led to the Central Coherence (CC) Hypothesis. According to this hypothesis autistic individuals show disturbances in the processing of global level information. In the present project the validation of the CC hypothesis is investigated using event-related potentials. Autistic and developmentally normal children will be compared at visual tasks that are thought to measure global and local processing and at which autistic individuals usually show deficient performances. These tasks are the Navon task, in which stimuli at both local and global levels have to be processed and a face-recognition task, in which upright and inverted faces and objects are presented. It will be investigated whether deficient performances of autistic individuals at these tasks can be replicated. Also it will be investigated whether autistic individuals show electrophysiological characteristics that are found to be related with local and global information processing. Besides this, a hypothesis will be formulated and tested concerning a possible underlying deficit in the visual information processing of autistic individuals. It is generally accepted that spatial frequencies play an important part in the processing of visual information. There is evidence that low spatial frequencies are specifically important in the processing of the stimuli in the Navon task and faces. Following the finding of their deficient performances at these tasks, it is hypothesized that autistic individuals may have deficient processing of low spatial frequencies. To test this hypothesis versions of the Navon task and the face-recognition task will be filtered with low-pass and high-pass filters. It is hypothesized that the differences between groups in behavioral performances and electrophysiological characteristics will be maximal at low-pass filtered stimuli and minimal at high-pass filtered stimuli. To asses whether potential deficits in the processing of low spatial frequencies exist at an early sensory level, ERPs will be measured in a task with low and high spatial frequency gratings. Source models will be determined to achieve more insight into the question whether there is a common source in the brain for the differences in the ERP activity of autistic individuals.