Surveys of freshwater mussel populations are used frequently to inform conservation decisions by providing information about the status and distribution of species. It is generally accepted that not all mussels or species are collected during surveys, and incomplete detection of individuals and species can bias data and can affect inferences. However, considerably less attention has been given to the potential effects of species misidentification. To evaluate the prevalence of and potential reasons for species misidentification, we conducted a laboratory-based identification exercise and quantified the relationships between mussel species characteristics, observer experience, and misidentification rate. We estimated that misidentification was fairly common, with rates averaging 27% across all species and ranging from 0 to 56%, and was related to mussel shell characteristics and observer experience. Most notably, species with shell texturing were 6.09× less likely than smooth-shelled species to be misidentified. Misidentification rates declined with observer experience, but for many species the risk of misidentification averaged >10% even for observers with moderate levels of experience (5–6 y). In addition, misidentification rates among observers showed substantial variability after controlling for experience. Our results suggest that species misidentification may be common in field surveys of freshwater mussels and could potentially bias estimates of population status and trends. Misidentification rates possibly could be reduced through use of regional workshops, testing and certification programs, and the availability of archived specimens and tissue samples in museum collections.