Although tuberculosis (TB) screening of immigrants has been conducted for over 50 yr in many industrialized countries, its cost- effectiveness has never been evaluated. We prospectively compared the yield and cost-effectiveness of two immigrant TB screening programs, using close-contact investigation and passive case detection. Study subjects included all immigration applicants undergoing radiographic screening, already arrived immigrants requiring surveillance for inactive TB, and close contacts of active cases resident in Montreal, Quebec, Canada, who were referred from June 1996 to June 1997 to the Montreal Chest Institute (MCI), a referral center specializing in respiratory diseases. For all subjects seen, demographic data, investigations, diagnoses, and therapy were abstracted from administrative data bases and medical charts. Estimated costs of detecting and treating each prevalent active case and preventing future active cases, based on federal and provincial health reimbursement schedules, were compared with the costs for passively diagnosed cases of active TB. Over a period of 1 yr, the three programs detected 27 cases of prevalent active TB and prevented 14 future cases. As compared with passive case detection, close-contact investigation resulted in net savings of $815 for each prevalent active case detected and treated and of $2,186 for each future active case prevented. The incremental cost to treat each case of prevalent active TB was $39,409 for applicant screening and $24,225 for surveillance, and the cost of preventing each case was $33,275 for applicants and $65,126 for surveillance. Close-contact investigation was highly cost effective and resulted in net savings. Immigrant applicant screening and surveillance programs had a significant impact but were much less cost effective, in large part because of substantial operational problems.