The mouse respiratory model is being used increasingly to study the pathogenesis and immunology of Bordetella pertussis infection. Two methods of inoculation, aerosol and intranasal, are routinely used to establish the infection. We compared the two methods of inoculation for reproducibility of infection using quantitative lung cultures and distribution of infection with [35S] methionine labeled bacteria and pulmonary histopathology. Ability to produce a respiratory infection intranasally was related to the inoculum volume; a minimum of 20 microliters was required although considerable variability remained. Lung bacterial counts in identically inoculated mice varied 1,000 fold following intranasal inoculation compared to only 5 fold following aerosol inoculation. Distribution of pulmonary 35S-labeled bacteria varied widely (right lung, 43-84%; left lung 16-57%) following intranasal in comparison to aerosol inoculation (right, 60-68%; left 32-40%). Finally, intranasal inoculation produced a scant, patchy, bronchopneumonia whereas diffuse pathology involving all pulmonary segments was seen following aerosol infection. Due to the superior reproducibility and predictable distribution of infection and pathology, aerosol inoculation is the method of choice for establishing the mouse model of pertussis respiratory infection.