OBJECTIVES: To determine the burden of pneumonia requiring hospitalization in infants and young children preventable by vaccination against Haemophilus influenzae type b (Hib). DESIGN: Vaccination centers in Santiago, Chile, were randomly selected to administer PRP-T, an Hib conjugate vaccine, combined with diphtheria-tetanus toxoids-pertussis (DTP) vaccine or DTP alone. SUBJECTS: Infants who received > or =2 doses of DTP or DTP and Hib conjugate vaccine combined. MAIN OUTCOME MEASURES: Pneumonia episodes leading to hospitalization accompanied by indicators of likely bacterial infection including radiologic evidence of alveolar consolidation or pleural effusion, an elevated erythrocyte sedimentation rate (> or =40 mm/h) or bronchial breath sounds on auscultation. RESULTS: In participants age 4 to 23 months, PRP-T reduced the incidence of pneumonia associated with alveolar consolidation or pleural effusion by 22% (95% confidence interval, -7 to 43) from 5.0 to 3.9 episodes per 1000 children per year. When the pneumonia case definition included any of the following, alveolar consolidation, pleural effusion, erythrocyte sedimentation rate > or =40 mm/h or bronchial breath sounds, PRP-T provided 26% protection (95% confidence interval, 7 to 44) and prevented 2.5 episodes per 1000 children per year. CONCLUSIONS: Hib vaccine provides substantial protection against nonbacteremic pneumonia, particularly those cases with alveolar consolidation, pleural effusion or other signs of likely bacterial infection. Hib vaccination prevented approximately 5 times as many nonbacteremic pneumonia cases in infants as meningitis cases, thus indicating that the largest part of the effect of Hib vaccination might be undetectable by routine culture methods.