For persons infected with Mycobacterium tuberculosis resistant to isoniazid (INH), rifampin is recommended for the prevention of active disease. However, the adverse effects and acceptability of this preventive therapy are largely uncharacterized. We prospectively followed 157 high-school students exposed to, and probably infected with, M. tuberculosis strains resistant to INH. All 157 students were prescribed preventive therapy with rifampin (10 mg/kg up to 600 mg daily) for 24 wk. While receiving therapy, 41 (26%) reported one or more adverse effects; of these, 18 had therapy interrupted temporarily, two permanently. Four (2.5%) had alanine aminotransferase elevations greater than two times the upper limit of normal (range, 91 to 161 U/L); of these, one had therapy permanently stopped. Six (3.8%) self-discontinued therapy. No student was found to have active disease during the 2 yr of the study (exact 95% upper confidence limit, 2.2). We assumed that without preventive therapy, seven cases of tuberculosis would have occurred during these 2 yr. Therefore, we estimated that rifampin had a minimum protective effect of 56%. In conclusion, preventive therapy with rifampin was well tolerated and well accepted, and it appears effective in preventing active tuberculosis.