At the end of the indentured labor schemes in the early years of the 20th century, laborers on the Samoan plantations were returned home. Most were initially repatriated to centres in Rabaul in East New Britain, or the Duke of York Islands, lying between New Britain and New Ireland in the Bismarck Archipelago (Mühlhäusler 1978). From there they were taken back to their home areas unless involved in local labour schemes. Further isolation from other south-west Pacific varieties led to considerable influence from the Austronesian languages of New Britain and New Ireland, especially in the lexicon, but also in grammatical structures. Features of the grammar of the early pidgin are also likely to have been reinforced if similar to structures widely present in local languages.

As noted, Papua New Guinea is an area of great linguistic diversity. A survey by the Summer Institute of Linguistics (Grimes 1992) lists over 860 languages currently spoken in a population of 4–5 million. At the beginning of the 20th century, poor communication and contact were the rule, with traditional trading activities operating along a complex though limited network of contacts. The upsurge in activities from overseas missions, traders and administrators led to an acute need for a language of wider communication, and the newly formed pidgin of the Samoan plantations, now fairly widely known, fitted ideally. In the monolingual Samoan society, however, it was no longer of any use, and soon died out there. The development of New Guinea Pidgin English thus proceeded in German-occupied New Guinea, and as it stabilized and expanded, it came under two influences not present in other varieties in Solomon Islands and New Hebrides.

The first of these was the language of the colonial power, German. A number of lexical items of German origin were adopted, especially in certain lexical fields, such as those related to education, woodworking, agriculture and so on, where German missionaries were intimately involved with the local population. Perhaps of equal significance was the fact that the English-lexicon pidgin was now effectively removed from further contact with its lexifier language.

The second influence on the stabilizing pidgin on the north coast of Mainland New Guinea was a substratum of non-Austronesian or Papuan languages. The languages of the Central Pacific as well as New Hebrides and Solomon Islands are almost uniformly Austronesian, and Austronesian languages are also dominant in the islands to the north and east of mainland New Guinea (Manus, New Britain, New Ireland and Bougainville). However, in parts of these areas, and most of the New Guinea mainland, the typologically different Papuan languages are spoken beyond a number of coastal enclaves of Austronesian speakers. The early pidgins exhibited a number of features typical of Austronesian languages, which tend to be reinforced by Austronesian-speaking populations, but there was little pressure to maintain exotic syntactic distinctions in non-Austronesian speaking areas. A good example of this is the so-called predicate marker i, which accords with the grammars of many Austronesian languages, and is thus retained in the Tok Pisin in these areas, but is routinely ignored in many non-Austronesian-speaking areas. Reesink (1990) has shown that some substrate syntactic features such as switch reference patterns and subordination are reflected in parallel differences in the Tok Pisin spoken in the area.