Abstract: OBJECTIVE To study the effect of various surfactants and various preparing process on in vivo distribution of penciclovir-loaded microemulsions. METHODS Drop sizes, zeta potential and in vitro release and in vivo distribution of three penciclovir-loaded microemulsions were investigated. They are Cremophor EL microemulsion by self-emulsifying (PCV-ME6), Cremophor EL microemulsion by homogenizing (PCV-Meho), sodium deoxycholate microemulsion by self-emulsifying (PCV-Mede). RESULTS Drop sizes of all penciclovir-loaded microemulsions were smaller than 100 nm, zeta potential between -3 mV and 0 mV. In vitro release of all penciclovir-loaded microemulsions were delayed, but only PCV-Mede showed evident difference from that of penciclovir solution (PCV-S) with f2 less than 50. PCV-Mede obviously increased penciclovir concentration in the brain and decelerated the clearance of penciclovir from kidney with AUC 3.5 folds higher than PCV-S. AUCs of PCV-Mede in plasma and other tissues also were distinctly higher than those of PCV-S and other penciclovir-loaded microemulsions. CONCLUSION What surfactants we choose greatly affect in vivo disposition of penciclovir. The microstructure of PCV-Mede might have changed during dilution after iv administration. A reassembly could take place between lecithin and sodium deoxycholate due to their strong interaction, which resulted in formation of lipid vesicles. Penciclovir is then solubilized into hydrophilic core of lipid vesicles and its release and disposition were changed. The different in vivo disposition of microemulsions made by self-emulsifying and homogenizing （PCV-ME6 & PCV-Meho）likely depended on less surfactant content of the later which is related to bigger drop sizes.