The current practice of using INH for tuberculosis prevention is limited by the necessity for at least 6 months of therapy and the problem of INH-induced hepatitis, particularly in older individuals and those with chronic liver disease. Bacteriologic models suggest that, in their persistent form, tubercle bacilli are relatively resistant to INH but become more sensitive to other drugs. Similarly, animal models of latent tuberculosis have suggested that alternative, short-course combinations such as RIF/PZA may be effective, and clinical trials of that two-drug regimen are continuing. At the present time, 3 months of daily RIF, 2 months of RIF/PZA, and 3 months of rifabutin can be considered reasonable alternatives to INH in selected patients. Routine use of these agents in preference to INH cannot yet be endorsed, however, as the standard of care. Without highly effective vaccines for tuberculosis, an important strategy for breaking the cycle of tuberculosis transmission lies in inexpensive, convenient, and effective preventive therapy.