Abstract Diverse control measures can be applied to reduce tuberculosis infection risk in health-care facilities. Selecting optimal controls requires methods for predicting the dependence of infections risk on underlying parameters. A common model for infection risk only explicitly accounts for control by ventilation. This paper proposes a more complete model for evaluating tuberculosis infection control methods in health-care settings. An infection risk parameter is defined as the probable number of infectious droplet nuclei inhaled by all susceptible persons from a single infectious person. Algebraic model equations are presented for two exposure cases. In one, the susceptible and infectious persons are together in a well-mixed indoor environment; in the socond, the infectious person is in respiratory isolation. Model equations are used to explore many common tuberculosis control measures: identification, isolation and treatment of tuberculosis cases; surgical masks and treatment booths applied at the source; environmental controls such as ventilation, air filtration, and ultraviolet germicidal irradiation; and respiratory protection for susceptible persons. Experimental data are limited or lacking on some key variables, such as emissions of infectious droplet nuclei by contagious persons and air leakage rates from isolation rooms. Methods are outlined for collecting additional data.