This paper describes an automatic system for labelling intonational tune information based on the Rise/Fall/Connection model of intonation. The system is powerful in that it presupposes no prosodic knowledge of the utterance it is recognizing, and is capable of labelling all the intonational tune effects of English.