To evaluate the protective efficacy of polyribosylribitol phosphate (PRP) and polyribosylribitol phosphate—diphtheria toxoid (PRP-D) vaccines in children 18 to 59 months of age, we conducted a case-control study in Los Angeles (Calif) County between July 1, 1988, and July 31, 1989. Seventy-nine children with invasive Haemophilus influenzae type b disease 18 to 59 months of age were identified, and 212 controls were selected by random-digit telephone dialing methods. Cases and controls were stratified by age and month of disease onset of the case. Seventeen PRP vaccine failures and two PRP-D vaccine failures occurred more than 2 weeks after vaccination. The PRP vaccine was shown not to be effective (point estimate - 47%; 95% confidence interval, - 307% to 47%), but the PRP-D vaccine was 88% protective (95% confidence interval, 42% to 97%). Adjustment of the efficacy estimates for potential confounding variables did not change the results significantly. The PRP-D vaccine provided significantly better protection than the PRP vaccine against invasive H influenzae type b disease in this population.