Substantial evidence suggests that sleep facilitates memory consolidation: the process by which new and initially weak memories become strengthened and resistant to interference. For instance, sleep plays a crucial role in the consolidation of newly learned language, working to strengthen memory for new words exactly as they were learned, integrate new words with existing lexical knowledge, and generalise existing lexical knowledge to new exemplars. The benefits of sleep for declarative memory are well established in adults, and more recent studies have revealed comparable or enhanced benefits in typically developing children. However, there is heterogeneity in the extent to which children consolidate new knowledge which cannot yet be fully explained by existing theoretical models. This talk will present data that strives to better understand this variability in vocabulary consolidation, drawing on studies of typical and atypical development. Three experiments will be presented that charter the time course of different aspects of language learning in children with and without autism spectrum disorder, specifically examining how sleep difficulties might account for differences in long-term retention. From a more practical perspective, data will be presented from studies that manipulate the time between learning and sleep, showing that learning closer to bedtime may be beneficial particularly for younger and vulnerable populations.