Chronic infectious disease is frequently perceived in a unitary fashion from the cellular point of view; i.e. all cells in a chronically infected host harbor the infectious agent, and the kinetics and result of infection are essentially the same in each infected cell. An alternative hypothesis is that these diseases involve simultaneous interrelated cycles of chronic and acute infections between a limited number of cell types. Infection of one cell type may be chronic and latent, produce little cytopathology and only rarely lead to the release of infectious units. Infection of other cell type(s) may be more acute, cause rapid and extensive cytopathic effects and the release of large numbers of infectious units. Signs and symptoms of chronic disease in the host are often dominated by damage to the cell type that exhibits the more acute and cytopathic response to infection.Interrelated cycles of chronic and acute patterns of infection in different cell types have been most clearly delineated in HIV infection[1xHo, D.D. J. Clin. Invest. 1997; 99: 2565–2567Crossref | PubMedSee all References][1]. Chronically HIV-infected dendritic cells and other mononuclear phagocytes show little cytopathic effect, whereas the raging acute HIV infection of activated CD4+ T cells leads to their destruction. A somewhat similar, but less completely defined, pattern occurs in herpes simplex infections[2xRaguin, G. and Malkin, J.E. Ann. Med. Interne (Paris). 1997; 148: 530–533PubMedSee all References][2]. In this case, the occasional production of infectious virions by chronically infected neurons is associated with the more acute and cytopathic mucocutaneous infection of epithelial cells. In malaria caused by Plasmodium vivax, a symptomatic, cytolytic, intraerythrocytic infection is superimposed on an essentially asymptomatic chronic hepatic infection[3xKumar, V. et al. Acta Leprol. 1992; 8: 87–94PubMedSee all References][3].Whether a similar pattern of interrelated cycles of infection and response of different cell types is also involved in the pathogenesis of chronic bacterial diseases is unclear. Infection of macrophages by Mycobacterium leprae is apparently not highly cytopathic in lepromatous leprosy, and direct infection of Schwann cells appears to be at least partly responsible for peripheral nerve dysfunction in this disease[3xKumar, V. et al. Acta Leprol. 1992; 8: 87–94PubMedSee all References, 4xBirdi, T.J., D'Souza, S., and Antia, H. Microb. Pathog. 1997; 22: 181–185Crossref | PubMed | Scopus (2)See all References], but evidence for a differential response to infection by these two cell types is limited[5xShetty, V.P., Uplekar, M.W., and Antia, N.H. Acta Neuropathol. 1994; 88: 300–306Crossref | PubMed | Scopus (26)See all References][5]. Tuberculosis presents an even more speculative case for interrelated acute and chronic infections of different cell types, as the types involved have not been fully defined. However, the relative inability of mice to control the growth of Mycobacterium tuberculosis in the lung, as compared with in the liver and spleen[6xMedina, E. and North, R.J. J. Exp. Med. 1996; 183: 1045–1051Crossref | PubMed | Scopus (81)See all References, 7xMedina, E. and North, R.J. Immunology. 1998; 93: 270–274Crossref | PubMed | Scopus (166)See all References], is consistent with an indolent infection in a cell type localized to the lung (e.g. pneumocytes or alveolar macrophages) with the occasional release of viable bacteria leading to active new infections in resident and recruited inflammatory macrophages. In both these diseases, the role of interrelated cycles of infection in pathogenesis is likely to be amplified by, and be in addition to, that played by the hypersensitivity of the host to mycobacterial antigens.Recognition of the role that dual patterns of infection may play in pathogenesis of chronic diseases caused by intracellular bacteria is likely to suggest new approaches to their study and lead to a more refined view of chronic infectious disease at the cellular level.