Recent concerns of declining bee populations have highlighted the importance of monitoring wild bees, but bee community assessments are hampered by species complexes that are difficult to identify. Bees in the genus Melissodes are often considered challenging to identify to species, with two widespread North American species, M. agilis Cresson and M. trinodis Robertson, being particularly difficult due to similar morphology, geographic ranges, and preferred floral hosts. These two species exhibit characteristics of cryptic species complexes, raising the possibility that our current understanding of their taxonomy is incomplete. We conducted a study to clarify the species boundaries within this complex, and to test if geometric morphometrics could be used to differentiate its member taxa. We sequenced fragments of the mitochondrial COI gene from 112 M. agilis/trinodis specimens, and integrated them into a phylogeny based on published reference sequences of over 70 Melissodes species. We additionally landmarked forewing venation for 102 of these specimens, tested if forewing morphometrics was associated with sex and phylogenetic clade, and tested if forewing morphometrics could accurately assign specimens to their proper clade and sex. Phylogenetic reconstructions resulted in nearly all specimens being assigned to three primary clades, with one clade containing reference sequences for M. agilis and M. trinodis, and two clades appearing to be undetermined cryptic taxa. Forewing morphometrics differed between clades and sexes, and was able to assign specimens to their proper clade or sex with over 80% accuracy, although accuracy of classification to clade declined to between 33-93% after cross-validation. Our results suggest the existence of cryptic diversity within M. agilis and M. trinodis, and indicate that forewing morphometrics can characterize some of this diversity. M. agilis and M. trinodis may comprise a complex of 3-5 cryptic species, but whether these are described or undescribed species is unknown. Also unclear is the degree to which the potential cryptic species contribute to the economically important sunflower pollination services currently considered to be conducted by M. agilis and M. trinodis. We encourage additional study of this complex to determine the nature of this cryptic diversity and resolve the taxonomic questions this study has raised.