PROJECT SUMMARY A major outstanding issue in sensory neuroscience is how the brain develops its remarkable ability to use senses together in synergistic ways to improve perceptual and behavioral decisions. Because each sense transduces a different form of environmental energy, powerful enhancements are obtained by synthesizing signals pertaining to the same event across the sensory modalities. There is an inherent ambiguity, however, in determining which signals belong to the same event, and which to different events (“causal inference”, a type of matching problem). Perhaps as a consequence of this ambiguity, multisensory integration is not an innate feature of the brain; rather, it must be acquired from experience with cross-modal events in the environment. How the brain uses this experience to solve the causal inference problem in development is not known. I hypothesize that the neonatal brain engages a form of statistical learning for this purpose. The algorithm it uses is predicted to operate on a site-specific basis, so that different multisensory combinatoric rules can be developed at different loci depending on the specifics of sensory experience. I postulate that the same plastic mechanisms are in place in the adult, but operate with reduced capacity. To study these questions, multisensory experience of both neonatal and adult naïve animals will be precisely controlled via a special rearing facility and sensory exposure chamber, and a combination of electrophysiological and behavioral methods will be leveraged to examine the functional properties of the multisensory superior colliculus (SC) and associated cortex (the anterior ectosylvian sulcus, AES).