1. A total of 499 rural and 437 urban Bantu school children were examined for their diphtheria carrier rate and Schick immunity.2. Clinical diphtheria amongst the rural and urban Bantu was also investigated.3. The virulent C.diphtheriaecarrier rate was found to be 3·2% in the rural children and 1·8% in the urban with virulent/avirulent ratios of 1: 1·1 and 1: 2·5 respectively.4. The Schick-positive rate in children aged 6–17 years was found to be 8·0% in rural children and 13·7 % in urban.5. Clinical diphtheria was rarely encountered under rural conditions, but was more common in urban natives.6. The case mortality rate in eighty-nine clinical cases was 14·4%.7. 53% of the clinical cases occurred in the 0-5 years age group.8. 89% of the strains recovered belonged to themitistype. No intermediate strains were encountered.9. The reason for the infrequency of clinical diphtheria amongst the Bantu is discussed and various theories are reviewed.10. It is concluded that the infrequency of clinical diphtheria is partly due to the environment, but that there is also a racial factor. It is suggested that the racial factor lies in an ability to produce antitoxin quickly. There is not sufficient evidence in this investigation to show whether the racial factor is genetic, but in view of Turbott's work amongst the Maoris it is suggested that the racial factor in immunity to diphtheria may be genetic amongst the Bantu also.I have pleasure in acknowledging the continued interest of Dr E. H. Cluver, Director, and Dr G. Buchanan, Deputy-Director, of the South African Institute for Medical Research in the progress of this work. I also wish to acknowledge gratefully the help given me by Dr Prestwick, Dr Miller and Dr Xuma at Alexandra Township. I am indebted to the Administration, and in particular to Dr J. W. Stirling, Principal Medical Officer of Bechuanaland Protectorate Government, for permission to carry out the investigation at Kanye. To Dr Marcus of the Seventh Day Adventist Mission, Kanye, I am deeply indebted for help and hospitality, and to Chief Bathoën who assisted me in making contact with the requisite number of school children at Kanye. My thanks are also due to the Mother Superior and Sisters of the Holy Cross Mission, Alexandra Township, and the many Bantu school teachers who gave me access to the children in their charge. Mr Barnes (S.A.I.M.R.) very kindly carried out the statistical tests of Tables 1 and 2. The travelling expenses involved in this work were defrayed by a grant from the National Research Board.