Recent studies have indicated that type 1 T cell responses (potent interferon-gamma and cytolytic responses, with absence of interleukin-4 production) are important for protective immunity against mycobacteria. These observations suggest that assays of type 1 T cell responses may be useful as surrogate markers of protective immunity in the evaluation of new tuberculosis vaccines. To be useful as surrogate markers, immunologic assays must distinguish between vaccine recipients and control subjects in clinical trials. Previous studies have shown that bacille Calmette-Guérin (BCG) vaccination can induce human type 1 T cell responses, but randomized trials have not been done to determine whether measurement of these responses can distinguish between BCG recipients and control subjects. We have conducted a double-blind, placebo-controlled trial of intradermal vaccination with two different BCG strains. We compared the mean lymphoproliferative, cytotoxic, Th1 and Th2 cytokine, and antibody responses detected in BCG and placebo recipients. These studies demonstrated that significant increases in Mycobacterium-specific T cell proliferative responses and type 1 cytokine responses were induced by BCG when compared with results with a placebo. In addition, BCG induced significant increases in Mycobacterium-specific antibody responses with an isotype profile characteristic of a type 1 cytokine bias. T cell and antibody assays involving the use of mycobacterial whole cell lysates or live BCG were able to discriminate between BCG and placebo recipients better than were assays using mycobacterial culture filtrates. These studies provide important information for the development of immunologic assays that might be useful as surrogate markers of protective immunity in future trials of new tuberculosis vaccines.