Previous literature shows that both explicit and incidental exposure to novel words can boost vocabulary. By comparing stories and definitions as ways of presenting novel words to children, the present study was able to assess the effectiveness of context on word-learning. It was predicted that learning would be greater when words are presented explicitly (in a definition) than incidentally (in a story). A comparison was also drawn between vocal and ‘passive’ interactions during word-teaching, where it was predicted that vocalised taught words will be learnt significantly better than words which are not vocalised by participants. Twenty preschool children were presented with 4 novel words, which were counterbalanced across 4 teaching conditions: words were presented explicitly through definitions or incidentally through stories, and were either vocalised or not vocalised by participants in each context. Each condition was a different combination of context-type and vocalisation-type (Definition Vocal/Non-vocal, Story Vocal/Non-vocal). The children then played 5 games, 3 of which tested word comprehension and 2 tested word production abilities. These were again played one week later to track learning over time. Using a repeated measures design, no significant difference was found between explicit teaching and incidental learning at Time 1 or Time 2. However, a significant difference between vocalisation and ‘passive’ learning was found at both Time 1 and Time 2. There was no effect of time on learning, but there was a significant difference between conditions at both times of testing, where the Story Vocal condition was significantly better than other teaching conditions. This suggests that stories are important tools for word-learning, but only when target words are vocalised during teaching. These findings have major implications on teaching methods, which may be applied to a more naturalistic setting.