The reported experiments were designed to study the contribution of prosodic cues to syntactic boundary categorization. The results indicate (1) that it is possible to differentiate between syntactic boundaries on the basis of prosodic cues alone (2) that the best results are obtained when pre- and post-boundary information is combined with information about the boundary itself (the silent interval) (3) that a fairly good categorization may be based exclusively on pre-boundary cues and (4) that the silent interval appears to be the stronger cue when in conflict with other prosodic cues.