The ability of a virus population to colonize a novel host is predicted to depend on the equilibrium frequency of potential colonists (i.e., genotypes capable of infecting the novel host) in the source population. In this study, we investigated the determinants of the equilibrium frequency of potential colonists in the RNA bacteriophage 6. We isolated 40 spontaneous mutants capable of infecting a novel Pseudomonas syringae host and sequenced their host attachment genes to identify the responsible mutations. We observed 16 different mutations in the host attachment gene and used a new statistical approach to estimate that 39 additional mutations were missed by our screen. Phenotypic and fitness assays confirmed that the proximate mechanism underlying host range expansion was an increase in the ability to attach to the novel host and that acquisition of this ability most often imposed a cost for growth rate on two standard hosts. Considered in a population genetic framework, our data suggest that host range mutations should exist in phage populations at an equilibrium frequency (3 x 10(-4)) that exceeds the phage mutation rate by more than two orders of magnitude. Thus, colonization of novel hosts is unlikely to be limited by an inability to produce appropriate mutations.