The metapodials of extinct horses have long been regarded as one of the most useful skeletal elements to determine taxonomic identity. However, recent research on both extant and extinct horses has revealed the possibility for plasticity in metapodial morphology, leading to notable variability within taxa. This calls into question the reliability of metapodials in species identification, particularly for species identified from fragmentary remains. Here, we use ten measurements of metapodials from 203 specimens of four Pleistocene horse species from eastern Beringia to test whether there are significant differences in metapodial morphology that support the presence of multiple species. We then reconstruct the body masses for every specimen to assess the range in body size within each species and determine whether species differ significantly from one another in mean body mass. We find that that taxonomic groups are based largely on the overall size of the metapodial, and that all metapodial measurements are highly autocorrelated. We also find that mean body mass differs significantly among most, but not all, species. We suggest that metapodial measurements are unreliable taxonomic indicators for Beringian horses given evidence for plasticity in metapodial morphology and their clear reflection of differences in body mass. We recommend future studies use more reliable indicators of taxonomy to identify Beringian horse species, particularly from localities from which fossils of several species have been recovered.