Late in 1991, before the implementation of a national immunization program against Haemophilus influenzae type b (Hib) in the United Kingdom, we performed a 4-year follow-up of 120 children who in 1987 had been enrolled in an immunogenicity trial in which 60 of them (vaccinees) received an Hib conjugate vaccine (HbOC) at the same time as diphtheria-tetanus toxoid-pertussis vaccine at the ages of 3, 5 and 9 months. Sixty others (controls) received only diphtheria-tetanus toxoid-pertussis vaccine at the same ages and were not subsequently immunized against Hib. We investigated Hib pharyngeal colonization using the antiserum agar method and the concentrations of serum IgG antibody to the type b capsule by enzyme-linked immunosorbent assay. At 4 years of age the Hib colonization rates in vaccinees and controls were 8\% (5 of 60) and 5\% (3 of 60), respectively. The children colonized with Hib had greater serum anti-capsular IgG concentrations than did noncolonized children (P \textless 0.001), and colonized vaccinees tended to have higher concentrations than colonized controls (P = 0.053). Regardless of Hib colonization status vaccinees had greater antibody concentrations than controls (P \textless 0.001). Forty-nine percent of vaccinees had an antibody concentration \textgreater 1 microgram/ml. There was an inverse relationship between the Hib colony count on culture and the serum IgG concentration. These data indicate that the increase in serum antibody concentration after immunization with an Hib conjugate vaccine is sustained and that immunization primes for a booster response on exposure to Hib. There may be a relationship between previous Hib conjugate immunization and the density of Hib colonization in children.