Ninety-two laboratories in the USA submitted isolates of Streptococcus pneumoniae to a single laboratory for susceptibility testing. Overall, 64% of 4489 isolates were susceptible to penicillin, 24% were intermediate and 13% were resistant to penicillin, although susceptibilities varied depending on geographical region. Macrolide/azalide resistance varied from 4 to 30%, with some regions having macrolide/azalide resistance higher than penicillin resistance. Children 12 years of age were significantly more likely to be infected with a penicillin-resistant pneumococcus than were adolescents or adults. Isolates from the respiratory tract were more likely to be penicillin resistant and >50% of pneumococci from the ear were resistant to penicillin. Almost 25% of penicillin-susceptible isolates had cefaclor MICs 2.0 mg/L and 15% of penicillin-susceptible isolates had loracarbef MICs 2.0 mg/L. These isolates would be erroneously reported as susceptible using NCCLS guidelines, and this finding may explain the lack of clinical response in patients treated with these antibiotics. The predicted plasma concentrations of all cephalosporins tested exceeded the geometric mean MIC for at least 40% of the dosing interval for penicillin-susceptible S. pneumoniae; for penicillin-intermediate S. pneumoniae, only cefprozil (56%), cefuroxime (64%) and cefpodoxime (63%) reached >40% of time above the geometric mean MIC in the dosing interval. None of the cephalosporins evaluated achieved a substantial time above the geometric mean MIC during its dosing interval for fully penicillin-resistant S. pneumoniae.