OBJECTIVE: To study the association between intensity of exposure and mortality due to pertussis. METHODS: Overall 3233 historical records from the main fever hospital in Copenhagen were examined. Exposure status of the whooping cough cases was coded as primary, secondary or multiple cases according to information in the hospital records. Primary cases, infected outside the home, were presumably exposed less intensively to the infectious agent compared to secondary cases, who were infected in the family. Multiple cases were from families with several simultaneous cases, but no clear information on transmission between cases. RESULTS: Case fatality was strongly related to age, being highest among the infants. Other risk factors were sibship size, exposure status and calendar period. In a multivariate analysis, period, age, and exposure status remained significant. Compared to primary cases, secondary cases and multiple cases had a 2.8 [RR = 2.76 (1.37 - 5.56)] and a two-fold [(RR = 1.99 (1.33 - 2.96)] higher risk of dying. CONCLUSION: Intensity of exposure may be a major determinant of the severity of pertussis infection.