OBJECTIVE To determine whether choice of a first- versus third-generation cephalosporin as initial therapy for lower respiratory tract infections in hospitalized adults affects the course and duration of care, both of which may influence antimicrobial treatment cost. METHODS Retrospective analysis of discharge abstracts and hospital pharmacy records. METHODS Forty-eight US acute-care hospitals. METHODS One thousand ninety-two hospitalized adults (aged > 17 y) with principal diagnoses of lower respiratory tract infections (DRGs 79-80, 89-90). METHODS Cefazolin or ceftriaxone, given as sole antimicrobial therapy for at least one day. METHODS (1) The number of patients who received another parenteral antibiotic anytime prior to hospital discharge; (2) the number of days during which patients received any parenteral antibiotic while in the hospital; and (3) the number of days patients remained hospitalized following the start of antibiotic therapy. RESULTS Patients treated with cefazolin (n = 763) were more likely to receive another parenteral antibiotic while in the hospital (30.3 vs. 20.7 percent; p < 0.001) and received more total days of therapy (7.2 vs. 6.7 d; p < 0.05) than those treated with ceftriaxone (n = 329). Although the time to hospital discharge did not differ in the full sample (9.2 d for both groups), it was greater among those receiving cefazolin (8.6 vs. 8.0 d; p < 0.05) when patients with lengths of stay exceeding 24 days were excluded from both groups. CONCLUSIONS In addition to acquisition cost, differences in course and duration of care should be considered when determining the most cost-effective choice for antimicrobial therapy.