It is generally assumed that elimination of Mycobacterium tuberculosis from the tissues of the protected individual depends on destruction of the organisms inside macrophages which have been activated by lymphokines secreted by T lymphocytes. However there has still been no convincing demonstration of kill or even powerful inhibition of M. tuberculosis by human macrophages. In fact this universally accepted mechanism of immunity remains strictly theoretical, and based mostly on the ease with which murine macrophages can be activated to inhibit this organism. A second surprising gap in our knowledge is the relationship between this hypothetical protective mechanism, and the immunopathology (necrosis) which characterises tuberculosis in man. Are both due to macrophage activation, and if so, are they the result of the activity of the same, or different lymphokines? This paper describes recent evidence that in man gamma-interferon (γ-IFN), the best characterised of the macorphage-activating lymphokines, may be more relevant to the immunopathology than to protection. This appears to be due to its effects on the regulation of vitamin D 3 metabolites, and of release of cachectin (also known as tumour necrosis factor, TNF).