Viral infection can cause changes to a plant?s morphology or chemical composition that may alter its desirability to the insect vector. In combination with viral persistence in the vector, prevalence of the disease, and the spatial distribution of infected plants, vector preference for virus-infected or uninfected individuals can strongly influence the rate of disease spread in plant communities (McElhany et al. 1995). Although it is known that the presence of a virus in the host plant can lead to a change in feeding behavior and an increase in fecundity (Fereres et al. 1990, Fereres et al. 1989, Montllor and Gildow 1986), there is little evidence that vectors prefer infected hosts over uninfected hosts. In this study, I examined host preference of the grass-feeding aphids Rhopalosiphum maidis, R. padi and Sitobion avenae when offered a variety of grasses infected or uninfected with the PAV species of barley yellow dwarf virus (BYDV). The eight grass species used in the preference tests are commonly found in New York State and are known hosts of BYDV. Preference tests were carried out in cages, where adult aphids were allowed to choose among grass seedlings. To determine whether infection altered preference, aphids were offered either uninfected grass communities or communities that contained a mixture of infected and uninfected grass seedlings. Results from these trials indicated strong species-specific preferences by aphids for particular grasses. Although there was some indication that infection could alter preference, I found no evidence that infection caused a significant shift in preference ranking among grasses.