The most important feature of the fricatives is the devoicing of the so-called voiced fricatives. It is not always clear whether the devoicing is phonemic or just phonetic, nor whether the same cause underlies all instances of fricative-devoicing. For example the pronunciation of thither with an initial [θ] is probably a lexical difference, parallel with the pronunciation found in Scottish English and some American varieties. The pronunciation of president as though homophonous with precedent seems more like a process of devoicing, which is currently variable in New Zealand English. There may nevertheless be a lexical dimension to this devoicing: president, positive seem particularly susceptible to it. So far, studies of the phenomenon have not distinguished between phonetic devoicing and vowel-shortening, so that it is not always clear whether a phonemic distinction is being lost or not. Certainly, it seems to be true that there is more sibilant-devoicing than there is corresponding vowel-shortening.

In /stj/ and /str/ clusters we find complex assimilation taking place. In /stj/ clusters there is coalescent assimilation of the /tj/ to [ʧ], and the post-alveolar quality is then passed on to the /s/ to give [ʃʧ], frequently heard in words like student. In /str/ clusters, the very slight retroflection of the /r/ was originally passed to the whole of the cluster, giving something that we might transcribe as  (although this seems to imply greater retroflexion than is actually found), but this has been reinterpreted by younger speakers as [ʃtɹ], as in words like strange.

/θ/ and /ð/ in New Zealand are usually interdental fricatives rather than postdental fricatives. An apparently innovative dental variant of /s/ has been described in studies carried out in Auckland, but it is not yet clear whether this is a regionalism or how widespread it is. There is some loss of /θ/ in favor of /f/, but this is not yet a major tendency.