OBJECTIVE: Our goal was to evaluate the immunogenicity and safety of pertussis booster vaccination in children infected with HIV on highly active antiretroviral therapy (HAART). PATIENTS AND METHODS: HIV-infected children on stable HAART for > or = 3 months with plasma HIV-RNA concentrations of < 30,000 to 60,000 copies per mL who previously received > or = 4 doses of diphtheria-tetanus-pertussis (DTP)-containing vaccine were eligible. Diphtheria-tetanus-acellular pertussis (DTaP) vaccine was administered to subjects 2 to < 7 years old who had 4 previous DTP-containing vaccines, subjects 2 to < 7 years old who had > or = 5 previous DTP-containing vaccines and negative tetanus antibody, and subjects > or = 7 to < or = 13 years old who had negative tetanus antibody. Pertussis toxin and filamentous hemagglutinin antibodies were measured before and 8, 24, and 72 weeks after DTaP vaccine. RESULTS: Ninety-two subjects received DTaP vaccine and met criteria for analysis. Antibody concentrations were low at entry: pertussis toxin geometric mean concentration at 4.8 enzyme-linked immunosorbent assay units (EU) per mL and filamentous hemagglutinin geometric mean concentration at 4.1 EU/mL. Pertussis toxin and filamentous hemagglutinin geometric mean concentrations rose to 22.3 and 77.0 EU/mL, respectively, 8 weeks after the study DTaP vaccine. Antibody concentrations fell by 24 weeks after vaccination but remained higher than before vaccination. Predictors of response 8 weeks after DTaP vaccine included the concentration of homologous antibody, lower HIV-RNA level, and higher CD4 percentage at entry. One vaccinated subject experienced erythema and induration of > or = 25 mm. CONCLUSIONS: A DTaP vaccine booster was well tolerated by children on HAART and induced increases in antibodies. Antibody concentrations after vaccination were lower than those reported in populations uninfected by HIV. Although comparison among studies must be made with caution, these data suggest that children infected with HIV may be deficient in immunologic memory from previous DTP-containing vaccination and/or that immune reconstitution with HAART may be incomplete for pertussis antigens.