This article describes the case of a client who displayed some interesting patterns of realization of the target English phoneme /r/. These varied according to both distribution within the word, and style of utterance. We speculate as to the cause of some of these forms, and on possible therapy strategies.