The aim of the present study was to make a comparison of the in vitro release rate of diclofenac sodium (DS) from microemulsion (M) vehicles containing soybean oil, nonionic surfactants (Brij 58 and Span 80), and different alcohols (ethanol [E], isopropyl alcohol [I], and propanol [P]) as cosurfactant. The optimum surfactant:cosurfactant (S:CoS) weight ratios and microemulsion areas were detected by the aid of phase diagrams. Three microemulsion formulations were selected, and their physicochemical properties were examined for the pH, viscosity, and conductivity. According to the release rate of DS, M prepared with P showed the significantly highest flux value (0.059 +/- 0.018 mg/cm(2)/h) among all formulations (P < .05). The conductivity results showed that DS-loaded microemulsions have higher conductivity values (18.8-20.2 microsiemens/cm) than unloaded formulations (16.9-17.9 microsiemens/cm), and loading DS into the formulation had no negative effect on system stability. Moreover, viscosity measurements were examined as a function of shear rate, and Newtonian fluid characterization was observed for each microemulsion system. All formulations had appropriate observed pH values varying from 6.70 to 6.85 for topical application. A skin irritation study was performed with microemulsions on human volunteers, and no visible reaction was observed with any of the formulations. In conclusion, M prepared with P may be a more appropriate formulation than the other 2 formulations studied as drug carrier for topical application.