Listeners customize speech processing to accommodate talker-specific phonetic variation. For example, listeners modify established phonetic boundaries to incorporate a talker’s idiosyncratic productions. In addition to being marked by boundaries, phonetic categories exhibit a graded internal structure, with some members of the category considered better members than others. Here we examined whether sensitivity to talker-specific phonetic variation influences internal category structure. Two groups of listeners heard a talker produce /k/. Word-initial voice-onset-time (VOT) was manipulated such that one group heard the talker produce /k/ with shorter VOTs relative to the other. Listeners were then presented with a range of VOTs and asked to rate each for goodness as /k/. Results to date indicate that exposure during training robustly influences the range of VOTs considered the best exemplars of /k/, suggesting that accommodating talker-specific phonetic variation results in a comprehensive re-mapping of acoustic-phonetic space and is not limited to the boundary region.