Many plant pathogens are transmitted by generalist insects that move between host species, and vector preference and performance on different hosts affects pathogen transmission. Most studies of host–vector–pathogen interactions focus on a single host, however, and it is relatively unknown if viruses affect the host breadth of vectors. For example, although pea aphids (Acyrthosiphon pisum) often form tight associations with specific hosts, viruses transmitted by aphids can break down host defenses, potentially weakening these associations and increasing the host breadth of their vectors. Here, we assessed vector host preference and performance in an experiment with five pea aphid colonies with distinct host associations, five host species, and two viruses (Pea enation mosaic virus [PEMV] and Bean leafroll virus [BLRV]). Our prediction was that pea aphids would have higher performance on their inferior hosts and be less discriminating among hosts when those are infected, compared with noninfected hosts. Our results show that aphid performance, measured as the rate of increase on hosts, differed among hosts based on their host association, and relative performance among hosts was indeed altered by virus infection. Aphid preferences, measured using a choice bioassay in which aphids could settle on any of the five host species, also differed among aphid colonies but was not affected by virus infection status. We also did not find that infection status of hosts altered the performance of aphids on their preferred host. Furthermore, virus infection did not disrupt preference performance relationships for the aphids. Overall, our study suggests that host preference and performance are deeply ingrained in pea aphids, and that host associations may be robust to the presence of host-manipulating plant pathogens.