Gist extraction is the process of excerpting shared features from a pool of new items. The present study examined sleep and the consolidation of gist in 12-month-old infants using a deferred imitation paradigm. Sixty infants were randomly assigned to a nap, a no-nap or a baseline control condition. In the nap and no-nap conditions, infants watched demonstrations of the same target actions on three different hand puppets that shared some features. During a 4-h delay, infants in the nap condition took a naturally scheduled nap while infants in the no-nap condition naturally stayed awake. Afterwards, infants were exposed to a novel forth hand puppet that combined some of the features from the previously encountered puppets. Only those infants who took a nap after learning produced a significantly higher number of target actions than infants in the baseline control condition who had not seen any demonstrations of target actions. Infants in the nap condition also produced significantly more target actions than infants in the no-nap condition. Sleep appears to support the storage of gist, which aids infants in applying recently acquired knowledge to novel circumstances.