The nonlinear mixed effect model (NONMEM) method for analysis of population pharmacokinetics, was applied to evaluate the relative biological availability of 2 ofloxacin (Ofl) products. The data of randomized crossover study of 2 kinds of Ofl tablets in 12 healthy volunteers were analyzed by NONMEM, as well as standard approach. The relative biological availability evaluated by NONMEM (F = 97.3 +/- 5.0%) was close to that by standard approach (F = 98.2 +/- 3.6%). NONMEM was also valid in accurately evaluating the relative biological availability using sparse clinical data (F = 91.7 +/- 8.0%) which could not be analyzed by standard approach. Thus, the usefulness and advantages of NONMEM were ascertained for the evaluation of biological availability using observational or experimental data.