Autistic children perform much better on abstract reasoning than their apparent intelligence and adaptation level. Growing empirical evidence suggests that both specific perceptual superiorities and different learning mechanisms characterize autism. This could lead to a preference for visual thinking: autistic individuals may think more in pictures than with words and hypotheses as others do. The proposed research aims at understanding the development of reasoning abilities in children on the autism spectrum. We will employ computerized tasks and functional brain imaging techniques to investigate reasoning development from 6 to 14 years old in autism spectrum children versus typical children. We aim to understand (1) the developmental trajectory of reasoning abilities in autistic children, and (2) modulation of reasoning mechanisms in relation to problem complexity. The long-term goal of this research program is to develop a better understanding of autistics' preferred modes of reasoning as well as cognitive and neural underpinnings of autistic reasoning, in order to develop targeted and efficient teaching and learning methods suitable to autistic children.