The main conclusions of this study can be summarised as follows:

Firstly, the scope of the term ‘urban’ with regard to homeland populations should be expanded beyond proclaimed towns (as stipulated by the 1980 census) to include settlements which are (a) dependent on commuting to proclaimed towns for labour, shopping and other services, (ie peri‐urban) and (b) have populations of more than 5 000 people (ie semi‐urban).

Secondly, the inclusion of the peri‐urban and semi‐urban populations pushes up the urban Black population of South Africa in 1980 from 8,6 million (as calculated by Simkins) to 10,7 million, and the level of Black urbanisation in South Africa from 41,6 per cent to 51,8 per cent. Within the homelands the effect is to more than double the urbanisation level from 15,6 per cent to 41,4 per cent. (See Tables 2 & 3, and Figures 2‐5)

Thirdly, with regard to what will happen with the abolition of influx control, the author argues that the most influential studies in this area are undermined by a number of shaky assumptions: namely, that migrants will be the first and only people to move from homelands to ‘White’ urban areas; that all migrants will move; that migrants will bring their household members with them; that they will move mostly across homeland borders into ‘White’ South Africa; and as a result, that substantial numbers (2,3 million in Simkins’ projection, and 8,2 million in that by Mostert el al) will move. We suggest that a more detailed examination of semi‐urban and recently resettled people is likely to yield more fruitful results.