The outcome of isoniazid chemoprophylaxis for tuberculosis was assessed in refugees from Southeast Asia. From July 1979 through June 1982, 2,795 tuberculin-positive refugees were prescribed isoniazid chemoprophylaxis at the time of their resettlement in King County, Washington. Through December 1983, 19 cases of active tuberculosis had arisen in those refugees. The annual incidence varied between 2.04 and 2.86 cases per 1,000 receiving treatment during the 4 yr after initiation of chemoprophylaxis (cumulative incidence, 9.66 per 1,000). Seven of the cases of tuberculosis occurring after isoniazid chemoprophylaxis were pulmonary, 1 was pulmonary and extrapulmonary (lymphatic), and 11 were extrapulmonary alone (lymphatic, 6 cases; pleural, 4 cases; soft tissue, 1 case). Fifteen of the 19 cases were culture positive, among which 8 M. tuberculosis isolates were susceptible to isoniazid and 7 were resistant. A case-control analysis showed that poor compliance, defined as taking isoniazid for 3 months or less, was associated with a sixfold increase in risk for subsequent isoniazid-susceptible tuberculosis but with no increased risk for subsequent isoniazid-resistant disease. Thus, active tuberculosis continued to occur with a low but relatively constant incidence after prescription of isoniazid chemoprophylaxis to Southeast Asian refugees. In this setting, poor compliance with chemoprophylaxis predisposed to isoniazid-susceptible, but not to isoniazid-resistant tuberculosis. There is a need to improve the approach to prevention of tuberculosis in Southeast Asian refugees.