Generation of genetic diversity is a broad theme in interactions between organisms and their changing environments. Only by modulating genetic diversity can a species optimize the rate and, to a very limited extent, the direction of its evolution. This dissertation examines a system for generating diversity in a bacteriophage protein that specifies tropism for host bacteria. The protein, major tropism determinant (Mtd) is diversified by replacement of a short stretch of protein coding DNA with a mutated copy, in a reverse transcriptase-mediated process. The system, termed a diversity-generating retroelement (DGR), served as model for the discovery a large family of DGRs spread throughout the bacterial kingdom. Interestingly, many of these DGRs are not phage encoded, but are instead found in bacterial genomes. Mtd appears to adopt a C-type lectin fold, a novel scaffold for high frequency diversification. Intriguingly, individual phage proteins interact only weakly with their host bacteria, but instead an array of Mtds, located at the distal ends of six phage tait fibers, work together to create a strong multivalent interaction with the bacteria. This pattern, termed avidity, is also important in interactions between antibodies and antigens. Despite the rapid evolution of Mtd, the rest of the bacteriophage structure remains constant, and its DNA-containing shell, or capsid, conforms to a structure seen even in eukaryotic viruses. The targeting of diversity to a short stretch of Mtd allows the bacteriophage to maintain the overall integrity of its genome while heavily mutating only the portion subjected to rapidly shifting selective pressures. Yet there is a price to pay even for this supremely focused diversity. Levels of a reverse transcriptase (RT) limit DGR activity in nature. When overexpressed, this RT can diversify Mtd at 100% frequency, forcing the phages to pay a significant penalty in reduced fitness.