Sequential distributions of intonational pitch accents in a handlabelled, read speech corpus of Tuscan Italian are analysed in relation to (1) the global frequencies of accents; (2) evidence of consistent patterns of association between accents. A subset of syntactically controlled nominal phrases are examined to determine whether the fact of their constituting a coherent syntactic unit has any effect on their patterns of sequential tone distribution, i.e. whether they are associated with a given tune. Some evidence is found for patterns of association between accents, and between sequences of accents and our chosen type of nominal phrase.