The scarcity of sizeable specimens of normal oral mucosa for experimental purposes has hampered research on oral epithelium. Because large specimens of viable human vaginal mucosa are readily available and because vaginal and buccal epithelia are microscopically similar, vaginal mucosa has been used successfully to establish a human cyst model in experimental animals. The ultrastructure and distribution of keratin filaments in these epithelia are also similar, as is their permeability to water and a number of chemical substances. Therefore, if vaginal mucosa could be substituted for buccal mucosa it would expedite research on the epithelium of buccal mucosa. To strengthen further the concept that vaginal epithelium could replace buccal epithelium in certain experimental studies, the thickness of these epithelia, their patterns of surface keratinization, the presence or absence of intercellular lipid lamellae and their lipid contents were now compared. Thirty-three specimens of vaginal mucosa from postmenopausal women and 36 of buccal mucosa were investigated. To compare the thickness of the epithelial layers the number of cell layers in sections of 20 vaginal and 20 buccal mucosal specimens were counted in the three thickest and three thinnest regions of each specimen. Surface keratinization was evaluated on sections stained with the Picro-Mallory method. To demonstrate lipid lamellae two vaginal and two buccal mucosa specimens were examined electron microscopically after normal fixation and postfixation in ruthenium tetroxide. Following solvent extraction of 11 vaginal and 14 buccal epithelia, quantitative lipid analyses were performed using thin-layer chromatography. No statistically significant differences were found between the maximum and minimum number of epithelial cell layers. The patterns of surface keratinization and the distribution and appearance of the lipid lamellae in the intercellular spaces were similar. The lipid composition of the two epithelia corresponded, except for the cholesterol esters and glycosylceramides, which were higher in buccal epithelium. Ceramides for vaginal epithelium and triglycerides for buccal epithelium were not determined. Based on structural similarities, a similar lipid composition and earlier findings, it is concluded that vaginal epithelium can be used as a substitute for buccal epithelium in certain in vitro, and possibly for in vivo, studies.