The declining incidence of AIDS-related opportunistic diseases among people with HIV infection has shifted the focus of clinical management to prevention and treatment of comorbidities such as chronic liver disease. The increased risk of hepatitis C virus (HCV)-related advanced liver disease in people with HIV infection makes early HCV diagnosis a priority. To assess HCV prevalence and predictors of HIV/HCV coinfection, we have conducted a retrospective analysis of people enrolled in the CAESAR (Canada, Australia, Europe, South Africa) study, a multinational randomized placebo-controlled study of the addition of lamivudine to background antiretroviral therapy. The impact of HCV on HIV disease progression was also examined. Anti-HCV antibody testing on 1649 CAESAR study participants demonstrated a HIV/HCV coinfection prevalence of 16.1%, which varied from 1.9% in South Africa to 48.6% in Italy. The strongest predictor of HIV/HCV coinfection was HIV exposure category (P<0.0001), with odds ratios (ORs) compared to homosexual as follows: injecting drug use (IDU), 365 [95% confidence interval (CI): 179-742]; transfusion or blood products, 32.2 (95% CI: 15.2-67.6); homosexual and IDU, 22.9 (95% CI: 8.5-62.1). The prevalence of HIV/HCV was low (3.7%) among homosexual men without reported IDU. Other predictors of HIV/HCV coinfection were alanine aminotransferase (ALT), country of residence, ethnicity and stage of HIV disease. A history of IDU or ALT > or =40 U/L at baseline had a positive predictive value (PPV) of 35%, negative predictive value (NPV) of 96%, sensitivity of 82% and specificity of 71% for HIV/HCV coinfection. HIV disease progression was similar in HIV monoinfected and HIV/HCV coinfected patients. People with HIV and a history of IDU or elevated liver function tests should be targeted for HCV testing. The low prevalence of HIV/HCV coinfection among homosexual men without a history of IDU suggests low efficiency of sexual HCV transmission.