In a seminal study Simcock and Hayne (2002) showed that 3-year-olds were unable to use newly acquired words to describe a "magic" event experienced 6 or 12 months earlier. In the reference study the children's verbal recall was tested without props being present. Inspired by recent evidence, the original design was replicated, testing 33-and 39-month-olds (n = 180), but with props present at recall while controlling for potential online reasoning. The results revealed that the children did use newly acquired words to describe their preverbal memory. Thus, the present study shows that nonverbal memories can be verbalized if the recall setting provides a high level of contextual support, a finding relevant to researchers investigating the offset of childhood amnesia.