A randomized, double-blind trial comparing a diphtheria-tetanus-acellular pertussis vaccine (DTaP) (pertussis toxoid and filamentous hemagglutinin) with a whole-cell vaccine (DTwP) was conducted. A case-contact study was nested in the trial to estimate absolute efficacy. From 1990 through 1994, 4181 children were randomized to receive one of the vaccines at 2, 4, and 6 months. Severe adverse events were monitored weekly during two visits after vaccination. Fewer serious adverse events were observed after DTaP. Surveillance for cough illnesses persisting more than 7 days, in children under 15 years of age, was made by weekly home visits. Examining physicians, blind to vaccination status, took samples for culture and serologic testing. Pertussis was defined as 21 or more days of cough confirmed by culture, serology, or contact with a culture-confirmed person. Beginning 28 days after the third vaccine dose, the overall ratio of pertussis incidence in the DTaP group relative to the DTwP group (RRac/wc) was 1.54 (95% CI, 1.23-1.93). In children younger than 18 months of age, RRac/wc was 1.16 (95% CI, 0.77-1.73) and 1.76 (95% CI, 1.33-2.33) in children older than 18 months, which suggests a shorter duration of protection with the acellular vaccine (P = 0.090). Absolute efficacy estimates derived from the case-contact study confirmed the lower protection afforded by the acellular vaccine compared with the whole-cell vaccine: 31% (95% CI, 7-49) versus 55% against the protocol case definition, and 85% (95% CI, 66-93) versus 96% for the more severe WHO case definition. Although vaccination with DTaP provided a lower degree of protection than the highly effective DTwP, this difference was less prominent before 18 months of age, the customary age for a fourth dose. The safer DTaP vaccine may prove a valuable substitute for whole-cell vaccines when used in a schedule that includes a booster-dose.