The morbidity of elderly patients with methicillin-resistant Staphylococcus aureus (MRSA) infections is high. In view of this, the purpose of the present study was to establish pharmaceutical health care for such patients having the aim of "proper use of antimicrobial agents for MRSA infection based on pharmacokinetics and pharmacodynamics ". I Vancomycin (VCM) 1) VCM exhibited a clinical effect in elderly patients infected with MRSA at serum trough levels of >10 μg/mL or an area under the curve (AUC 0-24) of > 420 μg / mL・hr. 2) Univariate logistic regression analysis of pharmacokinetic parameters indicated that VCM clearance was significantly enhanced in patients with malignant tumors. 3) With the population pharmacokinetics parameter mean (PM) method, the correlation between the value of the creatinine clearance (CLCr) estimated by substituting Jaffe-converted serum creatinine in the Cockcroft-Gault equation and the actually measured CLCr was the closest. 4) Prediction using Bayesian-estimated patient-specific parameters was more accurate than prediction by the PM method. II Linezolid (LZD) 1) The findings suggested a causal relationship between high LZD levels and adverse effects and that careful monitoring of LZD is necessary in hemodialysis patients. 2) The platelet count and hemoglobin level decreased as the AUC 0-24 of LZD increased in patients with renal dysfunction, and the correlations were significant with r =-0.593 and r =-0.783, respectively. 3) LZD is present mainly as the free drug in the cerebrospinal fluid (CSF). The free LZD concentration in the CSF can be estimated using the free serum concentration at the trough level. These findings suggest that we should apply pharmacokinetic analysis in order to ensure that anti-MRSA agents are used properly with emphasis placed on safety and efficacy.