Pertussis vaccines vary in quality, safety, and efficacy according to the production strains of Bordetella pertussis, the method of manufacture, and quality control procedures. It is therefore not justifiable to combine information on the incidence, nature, and severity of reactions after all manufacturers' pertussis vaccines as if they were a single product. Attempts were made to collect information on all suspected cases of severe reactions that occurred after administration of about 15 million doses of Wellcome pertussis vaccines in the United Kingdom and Northern Ireland from 1964 to mid-1977. Altogether six deaths, six neurological reactions with sequelae, and 17 convulsions without sequelae were reported, but some were clearly not attributable to the vaccine, while, in other cases, the available information was inadequate for assessing the role of vaccination. Neurological disorders, similar to those reported in a few children after pertussis vaccination, occur unexpectedly in apparently healthy infants at the recommended age for immunisation, so chance association between vaccination and these events can be expected in some children. The Joint Committee on Vaccination and Immunisation has made several recommendations aimed at reducing severe reactions after pertussis vaccination. These include replacing plain vaccine with aluminium-adsorbed vaccine, but there is no clear evidence that the aluminium-adsorbed vaccine produces fewer reactions than the plain.There are difficulties enough in deciding the cause of events that occur after vaccination, since these reactions often occur naturally in children of vaccination age. The task is made even harder by the assumption that various manufacturers' vaccines are the same and the lack of information available to manufacturers about cases in which their vaccine has been implicated. Information on vaccines administered is entered on immunisation records cards; it should be used and referred to if reactions occur.