The dissertation examines the relationship between prosodic/rhythmic structures of oral speech and their effect on simultaneous interpreting under different constraints in spontaneous and read-aloud speech. The data were collected from acoustic analysis systems and subjected to statistical tests. The conclusions show that there were significant differences between groups and between individuals regarding their rhythmic, melodic, relational and contextual characteristics. Read-aloud speech was clearly more difficult to interpret than spontaneous discourse. The results were explained by theories of dynamic attention and accent structures, and important parallelisms with musical theories were also found.