Technical Abstract: Recently best linear unbiased predictors (BLUPs) were used to determine the probable host-range of two plant pathogens proposed for classical biological control of Russian thistle. BLUPs were derived with mixed model equations incorporating disease ratings and relationship matrices computed from genetic (DNA) distances among plant species related to Russian thistle. For both pathogens, BLUPs were found superior to least squares estimates of disease severity and were better at delimiting the probable host-range of the pathogens among related species. Although previous work focused on evaluating disease severity on related plant species, this methodology can be used with any biological weed control agent as long as the evaluation criterion is quantitative or can be transformed to quantitative values. However, biological control researchers may be unfamiliar with the methodology and are possibly uncomfortable with the approach, since BLUPs are possible even for species which are not directly tested. In this paper, the uses and advantages of BLUPs in host-range determination are clarified.