Mermithids are filiform in shape with a smooth cuticle often possessing two distinct crossed layers of spiral fibers. The head usually has six flat cephalic papillae and well-developed amphids. The body has six or eight longitudinal cords, and the digestive tract is similar to free-living nematodes only in the young preparasitic larvae. In parasitic and postparasitic larvae and adults, the esophagus does not connect with the midintestine (trophosome) which is filled with food reserves. The esophagus extends one-fourth to nine-tenths of the body length and is almost devoid of musculature. The anus is absent in females. Gonads are paired in both sexes; in the female the genital aperture occurs at midbody, and in the male at the posterior end of the body in the form of one or two spicules. Eggs are round or oval, and in some terrestrial species have processes for adherence. The preparasite, postparasite, and adult worms are free living (Rubzov, 1972).

Mermithid taxonomy is in a state of confusion for several reasons. First, entomologists untrained in the biology and preservation of nematodes have historically been the persons encountering mermithids either by dissection of hosts or by observations of natural emergence. As a result, the nematodes were frequently destroyed during preservation or were improperly preserved as postparasitic juveniles. Second, mermithids present, at best, a difficult taxonomic group with few measurable morphological characteristics. Most of the early descriptions were incomplete and have become inadequate as more species are described and a more refined knowledge of this group is required. Third, many species have been described without knowledge of host associations and from only a few specimens. In some instances descriptions have been made from single specimens without knowledge of the opposite sex. Curran (1981) reported that many quantitative characteristics are affected by the environment, and suggested that many ratios commonly used in mermithid taxonomy should be rejected. Curran concluded that descriptions based on a single or a few specimens do not provide a sound foundation for mermithid taxonomy. The use of large samples, all developmental stages, and both quantitative and qualitative characters may overcome this problem. Fourth, improper mounting techniques distort morphological characteristics. Curran and Hominick (1980) found that qualitative characters, particularly head shape and cuticle thickness, and quantitative characters were differentially altered.

As a result of these classification problems, even the taxonomy of the most extensively studied mermithid, Romanomermis culicivorax, remains in doubt. This mermithid was first determined to be an undescribed species of Romanomermis (W. R. Nickle, personal communication). Nickle (1972) subsequently determined the Louisiana isolate to be the same as Reesimermis nielseni from Wyoming. In a detailed study of this group, Ross and Smith (1976) resurrected the genus Romanomermis and described the Louisiana isolate as R. culicivorax. Recently, Curran and Hominick (1980) reported that the characteristics used to separate R. culicivorax from a morphologically similar species R. iyengari (Galloway and Brust, 1979) were fixation artifacts and that R. culicivorax is species inquirenda until interbreeding studies can be done.

Poinar (1979) lists 32 genera of mermithids and suggests that there are over 200 species that have been described with many more still awaiting characterization.