The Bacillus subtilis purine repressor, PurR, regulates many genes involved in purine metabolism. These genes contain a conserved 14-nucleotide inverted repeat (PurBox). Both pur operon and purA, which are regulated by PurR, have this inverted repeat with a 16- or 17-nucleotide spacer, respectively. Mutational studies have earlier shown that PurR binding is dependent on the PurBox of pur operon. In contrast, these studies failed to establish the importance of purA PurBox to PurR binding. To examine this inconsistency, we studied the effects of PurBox mutations both in vivo and in vitro. The data presented here indicate that purA PurBox has a similar role as pur operon PurBox in PurR binding. In addition, our data suggest that the previously proposed classification of the two halves of the Purbox into weak and strong may need to be revised.