We constantly try to anticipate what we will see, hear and feel, in order to adapt our behavior to a continuously changing environment. How these predictive skills at lower perceptual levels influence acquisition of higher-order cognitive constructs such as language is still largely unknown. The current project aims to address this question within the domain of audiovisual speech perception. Typically, we not only hear but also see a speaker talk, and the presence of this visual information benefits speech recognition. This audiovisual benefit occurs as visual speech provides information that is complementary in its nature and often precedes auditory information (we see the lips close before we hear "p"). However, little is known about how audiovisual speech perception ability contributes to the development of speech sound (i.e., phonological) representations of words in the mental lexicon. This is the focus of the current project. To test the hypothesis that audiovisual prediction plays an important role in the development of phonological representations a longitudinal study highlighting the period of early literacy acquisition with children with and without a family history of dyslexia is proposed (Study A). Responses on behavioural audiovisual tasks reflect the combined result of perceptual and post-perceptual processes. In Study B, I will assess the dynamics of audiovisual perception directly by tracking the neural time-course of audiovisual perception for speech and non-speech, by examining event-related potentials. By comparing the time-course for speech to that of processing audiovisual non-speech events (i.e., seeing and hearing someone clap their hands), I will be able to determine whether any observed individual variation in the dynamics of audiovisual perception is limited to the language domain or is domain-general. The results will inform theories about the nature of phonological representations, and their development, as well as diagnostic and educational strategies for children at-risk of dyslexia.