We compared the susceptibilities of 100 Neisseria gonorrhoeae isolates obtained from January to June 2000 to a variety of antimicrobial agents, including various cephems, with the susceptibilities of 55 isolates obtained from May to December 1995. In 11 of the 15 cephems tested, the MIC50 (minimum inhibitory concentration for 50% of isolates) ratios, calculated by dividing the MIC50 for isolates from 2000 by that for isolates from 1995, showed 4-fold to 16-fold differences. The MIC90 ratios of cefodizime and cefditoren showed only a 1-fold and 2-fold difference, respectively. However, the MIC90 ratios of the remaining 13 cephems showed 4-fold to 64-fold differences. We selected the most favorable antibiotic regimens for the single-dose treatment of gonorrhea in Japan according to the antigonococcal activity and pharmacokinetics, based on the criterion (therapeutic time) proposed by Moran and Levine i.e., the regimen should produce a concentration in the serum or plasma at least four times that of the MIC90 value for at least 10 h. Of the 20 single-dose regimens evaluated, only ceftriaxone 1 g intravenously and cefodizime 1 g intravenously were found to have a therapeutic time of more than 10 h for the isolates from 2000. Our results indicated that the susceptibilities of the isolates from 2000 to all the cephems tested, except for cefditoren and cefodizime, decreased remarkably in comparison with the susceptibilities of the isolates from 1995, while parenteral ceftriaxone 1 g and cefodizime 1 g appeared to be the most favorable single-dose regimens for the treatment of gonococcal infections in Japan.