Aminoglycosides are a medically important class of antibiotics used to treat serious infections. Methylation of the ribosomal target is an emerging mechanism that produces a high level of resistance to all clinically available aminoglycosides for systemic therapy except streptomycin. ArmA was the first methyltransferase using this mechanism to be discovered in a clinical isolate. We demonstrate that ArmA methylates the N7 position of nucleotide G1405 in 16S rRNA. Methylation at this position is presumed to mediate cellular resistance by blocking aminoglycoside binding by ribosomes. To test this hypothesis, we measured the binding of gentamicin by 30S subunits. Under our conditions, we did not observe binding by ribosomes methylated by ArmA. Furthermore, the ArmA methylation reaction is specific for the 30S ribosomal subunit; neither 16S rRNA alone nor the 70S ribosome is a substrate for this reaction under our experimental conditions, implicating ribosomal proteins in substrate recognition. The biochemical characteristics of ArmA place it in the Agr family of methyltransferases, whose members are predominantly anti-suicide genes from Actinomycetes aminoglycoside producers. The discrepancy between the 30% GC content of armA and the >60% GC content of Actinomycetes, however, calls into question the origin of armA. We demonstrate that surprisingly, the natural promoter of armA from gram-negative Klebsiella pneumoniae was active in gram-positive Bacillus subtilis, suggesting that armA originated from a low-GC, gram-positive aminoglycoside-producing organism.