This chapter investigates, in a crosslinguistic perspective, the relationship between prosodic contours and direct and indirect reported speech (i.e. without or with deictic shift) in four typologically and genetically different Afroasiatic languages of the CorpAfroAs pilot corpus: Beja (Cushitic), Zaar (Chadic), Juba Arabic (Arabic based pidgin) and Modern Hebrew (Semitic). The descriptive tools and analysis of Genetti (2011) for direct speech report in Dolakha Newar (Tibeto-Burman) are used as a starting point and adapted to the annotation system of CorpAfroAs. Each language section investigates the prosodic cues and contours of direct speech reports, in relation to their quotative frame and their right and left contexts. As contradictory claims (e.g. Coulmas 1986 ; Klewitz & Couper-Kuhlen 1999 ; Jansen et al. 2001) have been made concerning the prosodic features of indirect reported speech, for example in English, the same prosodic features are also investigated for the three languages in our corpus which have indirect reported speech (Zaar, Juba Arabic and Hebrew). It is shown that speech reporting as a rhetorical strategy varies a lot from one language to another and is more frequent in the three unscripted languages of the sample. Even if speech reports show a wide range of prosodic behaviors, there are nonetheless clear tendencies that become apparent and which are related to various factors: speech report types, types of constituents of the quotative frame, genres, and typological features of the languages in question. A preliminary typology of the interface between prosody and speech reporting is proposed.