There is a growing concern with the vast individual differences in children’s vocabularies as they enter school. This study looked to determine the learning contexts and types of vocabulary instruction that are most effective for preschoolers by: (a) assessing how well preschoolers could learn new words inferentially from read-aloud stories compared to being told definitions explicitly and, (b) examining whether children’s own vocalization of new words affected their vocabulary acquisition. Furthermore, it aimed to determine the differential effects of these contexts on short- and long-term word retention, as well as on receptive and expressive vocabulary. In this within-subjects design, twenty preschoolers were exposed to four new words, each presented in one of four learning conditions: a read-aloud story, a read-aloud story involving novel word vocalization, a definition and a definition involving novel word vocalization. Accuracy and extent of word knowledge was measured immediately after learning and again a week later, using five assessment tasks (three of which measured comprehension abilities and two of which measured production abilities). Results indicate children’s immediate learning and long-term word retention was equally good after inferring word meaning from stories and hearing explicit definitions. Vocalization was found to primarily benefit children’s expressive vocabulary by enhancing short- and long-term word retention and boosting children’s inferential learning of words from stories. Vocalization was much less important for receptive vocabulary and learning from definitions. Vocalizing a new word encountered in a story was found to be the most effective overall learning context. The study provides a platform on which to develop more efficient and useful vocabulary teaching techniques for parents and nursery teachers.