Introduction. - Tuberculosis associated with HIV-infection in children makes the diagnosis of tuberculosis more complicated since it is already difficult to establish because clinically based in low-income countries, and worsens its outcome under treatment. We report our experience from the paediatric clinics of Bangui, Central African Republic. Patients and methods. - Our retrospective study analyzed 18- month - to 15- year-old children treated for tuberculosis from April 1998 to June 2000. Diagnosis and outcome data were abstracted from patient medical reports and we compared seropositive and seronegative patients. Results. - Globally, 284 cases have been analyzed. HIV-inf-ection rate was 25.7% (95% CI: 20.7- 31.2% ). Pulmonary tuberculosis and mixed forms rates were 94.4% (N=268). Extrapulmonary tuberculosis was essentially lymphadenopathies which have been restricted only to seronegative patients. Tuberculosis microbiological findings were significantly lower in seropositive patients compared with seronegative ones, for microscopy (8.2 vs 24.6% ) and for culture (35.6 vs 58.5% ) (P value< 0.05). On 28 seropositive and 72 seronegative children for which outcomes were registered, mortality rate was higher in seropositive than in seronegative patients (57.1 vs 19.4% respectively, P < 0.05). Conclusion. - The authors suggest that diagnosis of tuberculosis should be strengthened by blood or lymph node puncture culture particularly for HIV-infected children and that the treatment outcomes could be improved by diagnosis and treatment of other opportunistic infections. Introduction. - Tuberculosis associated with HIV-infection in children makes the diagnosis of tuberculosis more complicated because it is already difficult to establish due clinically based in low-income countries, and worsens its outcome under treatment. We report our experience from the pediatric clinics of Bangui, Central African Republic. Patients and methods. Our retrospective study analyzed 18- month - to 15-year-old children treated for tuberculosis from April 1998 to June 2000. Diagnosis and outcome data were abstracted from patient medical reports and we compared seropositive HIV-inf-ection rate was 25.7% (95% CI: 20.7-31.1%). Pulmonary tuberculosis and mixed forms rates were 94.4% (N = 268). Extrapulmonary tuberculosis was essentially lymphadenopathies which have been restricted only to seronegative patients. Tuberculosis microbiological findings were significantly lower in seropositive patien ts compared with seronegative ones for microscopy (8.2 vs 24.6%) and for culture (35.6 vs 58.5%) (P value <0.05). On 28 seropositive and 72 seronegative children for which outcomes were registered, mortality rate was higher in seropositive than Conclusion. The authors suggest that diagnosis of tuberculosis should be strengthened by blood or lymph node puncture culture particularly for HIV-infected children and that the treatment outcomes could be improved by diagnosis and treatment of other opportunistic infections.