This study examined the effects of instruction on the development of the English preposition to in a quasi-experimental setting with intermediate-level Japanese learners of English. Grammar tests were utilized to measure prepositional accuracies to understand the general effects of the treatment sessions that focused on developing knowledge of the preposition to. These tests were also used to verify if the prototypical feature of to (i.e., indicating an endpoint) and polysemous functions (i.e., spatial, temporal, and abstract) were correlated with changes in accuracies before and after the treatment sessions. Results of this investigation revealed that at the initial stage of this experiment there was little evidence that the notion of the prototypical meaning of to in the targeted items was understood, nor were these learners cognizant of the polysemy of the preposition to across spatial, temporal, or abstract contexts. Allowing for lexical substitution for the preposition to with before or until in the grammar tests, more temporal items could be correctly answered indicating a distinction between cognitive factors that may influence processing of the context, and the shared semantic properties of to, before, and until. After the treatment sessions, all targeted items reached very high accuracies in the post-tests for spatial and temporal scenes showing that learning effects could be maintained beyond the period of treatment. Abstract usages, on the other hand, had a lower level of achievement with a higher rate of attrition after the treatment. The general pedagogical implications of these findings suggest that the prototypical features of prepositions should be taught taking into account polysemy and various other factors such as semantic choice, cognitive features, developmental stages and systematicity, as they may affect explicit learning, retention, and implicit usage of prepositions.