Recent studies in populations with a high prevalence of tuberculosis and HIV infection report that tuberculous pleurisy occurs in approximately 30% of patients with tuberculosis. However, the fraction of patients with tuberculosis who have tuberculous pleurisy is comparable in HIV-positive and HIV-negative individuals. It appears that tuberculous pleurisy mostly develops in patients with HIV who have CD4 counts above 200/microL. Primary tuberculous pleurisy is thought to occur as a result of a delayed hypersensitivity reaction to mycobacterial antigens. Recent studies highlight the way in which pleural cells become activated and produce cytokines as a response to mycobacteria. Intramacrophage and direct cytotoxic elimination of mycobacteria, granuloma formation, and fibrosis are the main facets of this reaction. Many studies have investigated the usefulness of measuring different parameters in pleural fluid for an early diagnosis of tuberculous pleurisy. It has been shown that the most useful diagnostic tests are the levels of adenosine deaminase and interferon gamma in the pleural fluid. Elevation of either of these compounds in lymphocytic pleural effusions is virtually diagnostic of tuberculous pleurisy. Although theoretically, detection of mycobacterial DNA in the pleural fluid by the polymerase chain reaction would appear to be useful, the usefulness of this test still needs further demonstration. Patients with tuberculous pleurisy must receive antituberculous treatment. The current recommendation for immunocompetent patients is a 6-month regimen of isoniazid and rifampin supplemented in the first 2 months by pyrazinamide. HIV-infected patients should be treated with this same regimen for a longer time period. Serial thoracentesis or corticosteroid treatment are not warranted for the majority of patients.