Acellular pertussis vaccines were introduced nation-wide in Sweden in 1996, 17 years after the withdrawal of whole-cell pertussis vaccine from the childhood immunisation schedule. We report national data on age specific incidence of culture-confirmed Bordetella pertussis for 1986-2000, and clinical follow-up for 3 years (October 1997-September 2000) in children born in 1996-2000 and from children born in 1993-1994 who had participated in a trial of pertussis vaccines. The annual incidence of culture-confirmed B. pertussis was 89-150 per 100,000 before introduction of acellular pertussis vaccines and has dropped to 17-26 per 100,000. The data suggest that unimmunised infants and children who have received only one dose of pertussis vaccine were provided some protection. The decline is most obvious from the second dose onwards and remained stable for 4-5 years after the third dose in the absence of any booster dose. The first signs of waning immunity were observed at 6-7 years of age in the trial cohort. The short-term benefits reflect high vaccination coverage and high initial efficacy. The full impact of the acellular pertussis vaccination programme in infants remains to be established.