In 1950 a whole-cell pertussis vaccine was introduced in Cape Town and was followed by a marked decline in reported whooping cough mortality and morbidity. This resulted in reduced awareness of whooping cough as a clinical problem and, in recent years, no routine diagnostic tests for Bordetella pertussis have been performed. An outbreak of whooping cough occurred in Cape Town between 1 June 1988, and 31 May 1989, with 292 children admitted to hospital for whooping cough during this period (hospital admission rate in children under 5 years of age= 187 per 100 000). In an investigation of 239 children attending four pre-primary schools in the city, the whooping cough attack rate was 33 per cent, while pertussis vaccine coverage was 95 per cent. In the latter part of the outbreak nasopharyngeal swabs and serology were performed in patients presenting to a children's hospital with suspected whooping cough. Bordetella pertussis was isolated from 3 out of 34 (9 per cent) children tested and the first isolate was serotyped as type 1,2,4. Available clinical and laboratory evidence indicated that the organism responsible for the outbreak was Bordetella pertussis. Coverage studies for pertussis vaccine in Cape Town indicated that between 81 and 93 per cent of children were fully immunized by 13 months of age. These findings suggest that, since its introduction, the whole-cell pertussis vaccine produced in South Africa has been highly effective in controlling whooping cough. However, it was not able to prevent a moderate scale outbreak, even in the presence of high vaccination levels.