This article investigates the phonological relevance of the nucleus of the intonation contour. It does so primarily by reviewing the conditions under which stress shift occurs in English and Dutch, to see if such rules are in any way sensitive to the feature [nuclear]. One Dutch rhythm rule is identified which must refer to the feature. An attempt is made to relate the findings to the distinction claimed to exist in British English between a simple (mononuclear) fall-rise and a compound (binuclear) fall-plus-rise. It will be argued that no such distinction exists in either English or Dutch. It is pointed out that the existence of phonological rules that refer to the final sentence accent of a tone group does not imply that nonfinal sentence accents are distinct from the final one in a semantic sense.