Word-initial dental fricatives are realized as stops very widely, practically all around the globe. This applies even more strongly to the voiced /ð/ than to the voiceless /θ/. [d] for /ð/, especially in function words, is the rule rather than the exception throughout most of the Caribbean, in the Pacific contact varieties, in Africa and Asia (in South Asia the stop tends to be dental rather than alveolar) as well as in some dialects in America (notably AAVE, Nfl dE, and CajE) and Britain (IrE, BrC); in other British and American dialects it may also occur (but tends to be stigmatized). The distribution of [t] for /θ/ is very similar. Affricate realizations, on the other hand ([tθ] for /t/, [dð] for /d/), are fairly restricted, being possible in some American and Caribbean varieties and in Ghana. In word-central position the voiced dental fricative may be labialized, i.e. replaced by [v], although this is relatively rare (reported from a few British, American, and Caribbean varieties and Maori English). Word-final [f] for /θ/ may come up in AAVE, BahE, BrC, Ghana, SgE, and a few more British, American, Caribbean and antipodean dialects.

The voicing of word-initial fricatives is characteristic of south-western England but exceptional elsewhere (some evidence is provided for southern Wales, Newfoundland, and the Bahamas). The replacement of word-central labial fricatives by stops, e.g. riba ‘river’, is primarily characteristic of the Caribbean and reported as a rare possibility in a few African and Asian Englishes.

A voiceless velar fricative exists in ScE, WelE, and Northern Ireland, as well as, mostly in borrowings, WhSAfE and CFE.

The deletion of word-initial /h/ occurs variably in England, Wales, Nfl dE, IndE, MalE, AusE, NZE, and some forms of SAfE; typically it is associated with lower sociolinguistic status and informality. This feature occurs most regularly in the Caribbean (JamE/C, TobC, SurC, and elsewhere) and in the Pacifi c Pidgins, as well as, not surprisingly (given the possibility of transfer from French), CajE. The opposite process, /h/-insertion as in haxe ‘axe’, is even less widespread and also largely restricted to the Caribbean (with very few possible exceptions). In word-initial /hj/-clusters /h/ may be deleted in a fairly widely scattered array of varieties: This is reported as occurring regularly in East Anglia, New York City, urban SAmE, CajE, and CamE, and as a conditioned possibility in a few more dialects in Britain, America, the Caribbean, and South Africa. In South Africa, India and Pakistan /h/ may be voiced or murmured, according to our correspondents.

Further replacement processes have been observed primarily in specific African and Asian varieties, e.g. [s, z] for /ʃ , Ʒ/ , or the substitution of palatal fricatives for affricates. These processes seem to be more restricted and results of language contact.