A differential system for pathogen strains is the cornerstone to initiate a breeding program for disease resistance. The objective of this research was to develop a set of differential hosts for Bean pod mottle virus (BPMV). Seven BPMV isolates were used: one member of subgroup I (Ghabrial et al. 2001), three members of subgroup II, two reassortants of these subgroups, and one local isolate from Arkansas. Twenty-five cultivars of Glycine max, 114 lines of G. soja, seven accessions of G. tomentella, six genotypes of cowpea (Vigna sp.), and five bean cultivars (Phaseolus spp.) were screened as potential differential hosts. However, no resistance, meaning reduced or absence of mosaic symptoms, was found after five repeated experiments over time and under different environmental conditions. Nonetheless, PI464928 and PI464889A (G. soja accessions) developed stem tip necrosis to the reassortant isolates, and expressed mosaic symptoms to the isolates of subgroup I, II and AR. It was, therefore, possible to separate reassortants from the other groups of isolates with either of these two G. soja accessions. In conclusion, although it was not possible to classify the isolates into physiological races, the variation among BPMV isolates allowed the separation of severe reassortants from the milder subgroups of isolates.