This discussion covers the chemical structure of cervical mucus, accepted thoeries of its ultrastructure, and the author's data from the scanning electron microscope. A theory of the ultrastructure of cervical mucus must explain how it can be viscous and hostile to sperm for most of the cycle, but elastic and arranged parallel to their upward migration at ovulation. Cervical mucus is a hydrogel with 2%-12% solid phase composed of glucoproteins, probably meshed noncovalently into protein chains, with oligosaccharide side chains ending in sialic acid. A popular thoery generated by nuclear magnetic resonnance studies suggests that there may be sheaves of fibers arranged into micelles, with transverse fibers forming a netwrok that enlarges at ovulation. The light microscope is useless for studying mucus structure, but transmission electron microscopes have tentatively verified this hypothesis. The author's work with the scanning electron microscope showed a tangled web of filaments approximately 500-750 Angstroms, 1000-1500 or 300-6000 Angstroms thick. Usually the margin of the specimen appeared thin, like a spider web, but the center appeared thick with open channels, like the skeleton of a sponge, with a secondary network at the level of the oblique and transverse fibers.