T he human immune response to Mycobacterium tuberculosis is not well characterized. To better understand the cellular immune response to tuberculosis, a human mononuclear phagocyte culture system using a low-infecting inoculum of M. tuberculosis to mimic in vivo conditions was developed. Using this system, monocytes treated with IFNgamma/TNFalpha/ calcitriol (CytD) were permissive for the growth of virulent M. tuberculosis. In the presence of iron, however, these monocytes suppressed the growth of M. tuberculosis. The enhanced permissiveness of CytD-preincubated monocytes was found to be due to TNFalpha, however, the ability of iron to suppress M. tuberculosis growth also required preincubation with TNFalpha. Iron-mediated growth suppression was correlated with selective suppression of TNFalpha release from infected monocytes. In addition, removal of TNFalpha from CytD-treated monocytes 2 d after infection mimicked the suppressive effect of iron, suggesting that iron may also be decreasing monocyte sensitivity to exogenously added TNFalpha. In the absence of iron, permissive, CytD-treated monocytes formed large infected cellular aggregates. With iron treatment, aggregation was suppressed, suggesting that the iron-suppressive effect on M. tuberculosis growth may be related to suppression of monocyte aggregation and diminished cell-to-cell spread of M. tuberculosis. The results of this study indicate that TNFalpha preincubation is required for human monocytes to exert an iron-mediated suppressive effect on M. tuberculosis growth. In the absence of iron, however, the continued presence of TNFalpha has a growth-promoting effect on M. tuberculosis in human monocytes. Iron may be an important early modulator of M. tuberculosis growth via its effects on TNFalpha.