An increasing prevalence of methicillin-resistant Staphylococcus (S.) aureus (MRSA) has a serious therapeutic problem, and accurate methods to detect such strains are needed. We studied the antimicrobial susceptibility of S. aureus to ceftizoxime, in comparison with those to four other cephems (cefazolin, cefoxitin, latamoxef and cefmenoxime), by broth microdilutions and disk susceptibility tests, and also evaluated whether the reagents, in replace of penicillinase-resistant penicillins (PRPs), could discriminate between the strains of MRSA and those susceptible to PRPs (MSSA). A total of 651 clinical isolates of S. aureus were collected from six geographically different hospitals. All the strains collected were first classified into either MRSA (n = 329) or MSSA (n = 322) according to the interpretations of MRSA screening agar, minimum inhibitory concentrations (MICs) to oxacillin and methicillin (NCCLS M7-A2), and the presence or absence of mecA gene by polymerase chain reaction. In broth microdilution tests, the MICs of MRSA to ceftizoxime ranged > or = 64 micrograms/ml, whereas all the MSSA were at the concentration of < or = 16 micrograms/ml. The results of Showa disk diffusion tests highly correlated with those of MIC determinations. The distribution of inhibitory zone diameters to ceftizoxime were clearly divided into two groups; 99.2% (sensitivity) of MRSA had inhibitory zones of < or = 20 mm and 98.9% (specificity) of MSSA produced > or = 21 mm. It was concluded that the Showa ceftizoxime disk susceptibility test was useful and enough reliable to screen MRSA isolates in clinical laboratories.