Fluoroquinolone (FQ)-resistant bacteria pose a major global health threat. Unanalysed genomic data from thousands of sequenced microbes likely contain important hints regarding the evolution of FQ resistance, yet this information lies fallow. Here we analysed the co-occurrence patterns of quinolone resistance mutations in genes encoding the FQ drug targets DNA gyrase (gyrase) and topoisomerase IV (topo-IV) from 36,402 bacterial genomes, representing 10 Gram-positive and 10 Gram-negative species. For 19 species, the likeliest routes toward resistance mutations in both targets were determined, and for 5 species those mutations necessary and sufficient to predict FQ resistance were also determined. Target mutation hierarchy was fixed in all examined Gram-negative species, with gyrase being the primary and topo-IV the secondary quinolone target, as well as in six of nine Gram-positive species, with topo-IV being the primary and gyrase the secondary target. By contrast, in three Gram-positive species (Staphylococcus haemolyticus, Streptococcus pneumoniae and Streptococcus suis), under some conditions gyrase became the primary and topo-IV the secondary target. The path through individual resistance mutations varied by species. Both linear and branched paths were identified in Gram-positive and Gram-negative organisms alike. Finally, FQ resistance could be predicted based solely on target gene quinolone resistance mutations for Acinetobacter baumannii, Escherichia coli and Staphylococcus aureus, but not Klebsiella pneumoniae or Pseudomonas aeruginosa. These findings have important implications both for sequence-based diagnostics and for understanding the emergence of FQ resistance.