Human cognition is supported, in part, by adaptive changes in perception that occur as a result of experience. This dissertation proposes a mechanism by which perception can adaptively change to reflect the structure of the environment. Environmental structure produces statistical regularities in sensory input. Starting very early in infancy and continuing across the lifespan, humans have the ability to pick up on these statistical regularities through a process called statistical learning. This dissertation presents empirical findings that statistical learning occurs through a dynamic and mutually-influential relationship between perceptual and learning and memory systems, where perceptual processes support and constrain learning, and learning and memory systems, in turn, shape future perception. To this end, infants and adults participated in incidental learning paradigms. In these paradigms, participants experience sensory input endowed with novel environmental structure that can be learned through statistical regularities. After exposure, memory for the structure is tested. A number of experiments examined learning where the statistical information is held constant but the perceptual character of the information varied (e.g., whether the stimuli are auditory or visual or whether statistical information is based on perceptually varying exemplars). Using these methods, Chapters 2 and 3 present evidence that auditory and visual perceptual processing biases what is learned from novel statistical regularities. Thus, perceptual processing can affect statistical learning even though statistical information is held constant. Chapter 4 presents evidence for a dynamic influence of perception on statistical learning throughout development. Specifically, the results in this chapter demonstrate that the relationship of auditory and visual learning is different in infant and adult learners and that auditory statistical learning changes in infants aged eight to ten months. Finally, Chapter 5 presents evidence that changes in perception can occur through experience with environmental structure and points to the involvement of learning and memory systems in this process. Experiments in this chapter find that variable yet regular exposure with a novel object supports changes in object perception. A combined eye-tracking/functional magnetic resonance imaging (fMRI) methods approach highlights the involvement of the medial temporal lobe, an important learning and memory system, in integrating across successive experiences to support changes in object perception. Taken together, this dissertation presents empirical evidence that during experience with novel statistical regularities, perceptual processing affects learning, and learning and memory systems can affect perception. The interrelationship of perceptual and learning and memory systems could act as a dynamic mechanism supporting adaptive changes in perception across the lifespan. Implications for the fields of Developmental Psychology, Cognitive Psychology and Cognitive Neuroscience are explored.