Ribonucleotide reductase (RNR) is an enzyme family required for the de novo synthesis and fidelity of DNA in every free-living organism. The enzyme is found in every ecological niche and is responsible for converting the precursor building blocks of RNA (ribonucleotides) to the precursor building blocks of DNA (2′-deoxyribonucleotides). RNRs are necessarily diverse at the primary sequence level to adapt to the diversity of their host organisms, however they maintain a conserved fold about the active site and utilize a conserved radical-based mechanism for catalysis. This thesis uses bioinformatic tools (sequence similarity networks, phylogenetic inference, ancestor sequence reconstruction, statistical coupling analysis) combined with structural characterizations (small angle X-ray scattering, cryogenic electron microscopy) to predict function from primary sequence in the metabolically essential family of RNR. The work presented in this thesis determines new insights into the evolution of allostery in RNRs while providing a general approach for targeted study of diverse enzyme families.