To determine whether unionids contain surface-attached endosymbiotic bacteria, ciliates, or fungi, we used scanning electron microscopy to examine the epithelial surface of various organs within the digestive systems and mantle cavity of temperate river and lake unionids on a seasonal basis. We also cultured material removed from the lumen of these same organs and from the mantle cavity to detect cellobiose-, cellulose-, and chitin- degrading microbes. No true endosymbiotic fauna were observed attached to the surface of the digestive or mantle tissues of any species of unionid. Microbial growth on cellulose or chitin bacteriological media varied by season and habitat, but not by unionid species or source of the isolate. Lake unionids did not contain microbes capable of digesting cellulose or chitin, whereas unionids from the river site did in March and August, but not in December. Since these cultured cellulose- and chitin-degrading bacteria were never found attached to any unionid tissues, they appeared to be transient forms, not true endosymbionts. Microbes capable of digesting cellobiose were found in every unionid collected in all seasons and habitats, but again, no microbes were directly observed attached to unionid tissues. If unionids, like most other vertebrates, lack digestive enzymes required to initiate primary bond breakage, then the lack of cellulolytic and chitinolytic endosymbionts would affect the ability to utilize cellulose or chitin foods. Thus, in captivity dry feeds based on corn, soybeans, or nauplii should be pre-digested to ensure maximum absorption of nutrients by unionids. The lack of celluloytic or chitinolytic endosymbionts should not affect relocation success, though the seasonal role of transient microbes in unionid nutrition requires further investigation.