NrdR is a universal transcriptional repressor of bacterial genes coding for ribonucleotide reductases (RNRs), essential enzymes that provide DNA building blocks in all living cells. Despite its bacterial prevalence, the NrdR mechanism has been scarcely studied. We report the biochemical, biophysical, and bioinformatical characterization of NrdR and its binding sites from two major bacterial pathogens of the phylum Bacillota Listeria monocytogenes and Streptococcus pneumoniae. NrdR consists of a Zn-ribbon domain followed by an ATP-cone domain. We show that it forms tetramers that bind to DNA when loaded with ATP and dATP, but if loaded with only ATP, NrdR forms various oligomeric complexes unable to bind DNA. The DNA-binding site in L. monocytogenes is a pair of NrdR boxes separated by 15-16 bp, whereas in S. pneumoniae, the NrdR boxes are separated by unusually long spacers of 25-26 bp. This observation triggered a comprehensive binding study of four NrdRs from L. monocytogenes, S. pneumoniae, Escherichia coli, and Streptomyces coelicolor to a series of dsDNA fragments where the NrdR boxes were separated by 12-27 bp. The in vitro results were confirmed in vivo in E. coli and revealed that NrdR binds most efficiently when there is an integer number of DNA turns between the center of the two NrdR boxes. The study facilitates the prediction of NrdR binding sites in bacterial genomes and suggests that the NrdR mechanism is conserved throughout the bacterial domain. It sheds light on RNR regulation in Listeria and Streptococcus, and since NrdR does not occur in eukaryotes, opens a way to the development of novel antibiotics.