To examine the utility of the Pneumonia PORT prediction rule for evaluating the severity of community-acquired pneumonia (CAP) in Japan, 127 consecutive patients with CAP admitted during one-year period were recruited and stratified by the rule. The relationship between CAP severity determined by the rule and drug-resistant Streptococcus pneumoniae was also investigated. The number of the patients classified into each class was 20 (15.7%) in class I, 32 (25.2%) in class II, 32 (25.2%) in class III, 31 (24.4%) in class IV, and 12 (9.4%) in class V, respectively. Risk class-specific mortality rate was 0% in class I and II, 3.1% in class III, 3.2% in class IV, and 41.7% in class V, respectively. Twenty-seven patients (21.3%) were diagnosed as pneumococcal pneumonia. In terms of penicillin-susceptibility of Streptococcus pneumoniae, 19 strains (70.4%) were penicillin-susceptible Streptococcus pneumoniae (PSSP). Five (18.5%) were penicillin-intermediate Streptococcus pneumoniae (PISP), and 3 (11.1%) were penicillin-resistant Streptococcus pneumoniae (PRSP). Highly resistant pneumococci with penicillin MIC > or = 4.0 micrograms/mL was not observed. For severity of drug-resistant pneumococcal pneumonia, 3 patients infected with PISP were classified as class IV, 2 with PISP were as class V, 2 with PRSP were as class IV, and 1 with PRSP was as class V. The patients with drug-resistant pneumococcal pneumonia were classified as high risk classes (class IV to V), however, the mortality rate was low (0%). In conclusion, the Pneumonia PORT prediction rule is effective for evaluating the severity of CAP in Japan, especially in classifying the class V patients who are related to high mortality. The mortality rate of the patients with drug-resistant pneumococcal pneumonia was low.