The duration of time that serum levels are above the minimum inhibitory concentration (MIC; T >MIC) seems to be an important pharmacodynamic parameter for beta-lactams. The aim of this study was to evaluate the bactericidal activity of cefodizime and ceftriaxone in a pharmacokinetic model mimicking the concentrations in bronchial mucus and in serum (total and free) obtained at 2, 4, 8, 12 and 24 h, after 1 g i.m. administration once daily. The species investigated were respiratory pathogens (1 strain of Staphylococcus aureus, 2 strains of Streptococcus pneumoniae, 1 strain b-lactamase negative and 1 strain beta-lactamase positive of Haemophilus influenzae, 1 strain of Escherichia coli and 1 strain of Klebsiella pneumoniae); MIC50s of the chosen strains were reported. In this in vitro model the concentrations (serum and bronchial mucus) for both antibiotics are generally at or above the MIC values of the tested strains until 24 hours. The killing curve showed rapid killing for both antibiotics: 99.9% killing (a 3-log reduction in growth) within 6 to 8 h, depending upon the microorganism tested. There was no significant difference in the log kill between cefodizime and ceftriaxone. These data confirm that T >MIC for beta-lactams is the pharmacodynamic parameter which best correlates with bactericidal efficacy. On the basis of the killing curve determined for cefodizime versus ceftriaxone at concentrations that these antibiotics can reach during therapy with 1 g i.m. once daily we expect reasonable clinical efficacy with monoadministration of cefodizime as well as for ceftriaxone in respiratory tract infections.