Respiratory tract infections are treated empirically. Treatment is based on the likely pathogens and their antibiotic susceptibility. The most common respiratory tract pathogen is Streptococcus pneumoniae. In the United States, approximately 25% to 30% of S. pneumoniae are resistant to erythromycin and other macrolides. There are two mechanisms of resistance: ribosomal methylation that causes high-level resistance, and an efflux pump that causes low-level resistance. Macrolides are ineffective in animal models that use pneumococcal isolates with the methylase- or efflux-mediated resistance mechanisms. There are many case reports that describe clinical failure and isolation of a macrolide-resistant pneumococcus while a patient receives macrolide treatment. Two recent studies that included macrolide-susceptible and macrolide-resistant pneumococci showed that breakthrough bacteremia in patients receiving macrolide treatment occurred only with macrolide-resistant isolates. Study of bacteremic disease ensures the pathogenic role of the pneumococcus; however, it underestimates the true clinical impact of macrolide resistance.