OBJECTIVES: To investigate the presence of hepatitis B virus (HBV) DNA and hepatitis C virus (HCV) RNA in HIV-infected patients initiating antiretroviral therapy in Cameroon. METHODS: Baseline blood samples from 169 patients were tested retrospectively for hepatitis B surface antigens (HBsAg), anti-hepatitis B core (anti-HBc), anti-HCV and - if HBsAg or anti-HCV result was positive or indeterminate - for HBV DNA or HCV RNA, respectively, using the Cobas Ampliprep/Cobas TaqMan quantitative assay (Roche Diagnostics GmbH, Mannheim, Germany). RESULTS: HBV DNA was detected in 14 of the 18 patients with positive or indeterminate HBsAg results [8.3% of the total study population, 95% confidence interval (CI) 4.6-13.5]. The median HBV viral load was 2.47 x 10(7) IU/mL [interquartile range (IQR) 3680-1.59 x 10(8); range 270 to >2.2 x 10(8)]. Twenty-one patients (12.4%, 95% CI 7.9-18.4) were found with HCV RNA (all with positive HCV serology). The median HCV viral load was 928 000 IU/mL (IQR 178 400-2.06 x 10(6); range 640-5.5 x 10(6)). No patient was co-infected with HBV and HCV. In multivariate analysis, HCV co-infection was associated with greater age [ ≥ 45 years vs. <45 years, odds ratio (OR) 11.89, 95% CI 3.49-40.55, P<0.001] and abnormal serum alanine aminotransferase level [ ≥ 1.25 x upper limit of normal (ULN) vs. <1.25 x ULN, OR 7.81, 95% CI 1.54-39.66, P=0.01]; HBV co-infection was associated with abnormal serum aspartate aminotransferase level (OR 4.33, 95% CI 1.32-14.17, P=0.02). CONCLUSIONS: These high rates of active HBV and HCV co-infections in HIV-positive Cameroonian patients requiring antiretroviral therapy underline the need to promote: (i) screening for HBV and HCV before treatment initiation; (ii) accessibility to tenofovir (especially in HBV-endemic African countries); and (iii) accessibility to treatment for HBV and HCV infections.