Am J Clin Pathol 63:438-445, 1975. The effects of procedural variations on the activities of gentamicin, tobramycin, sisomicin, kanamycin, and amikacin in vitro were evaluated using 134 clinical isolates. In broth dilution studies, a change in assay medium from brain-heart infusion broth to Mueller-Hinton broth resulted in significant changes in minimal inhibitory concentrations in 36% (242 of 670) of assays. A change in the bacterial population size utilized in broth dilution studies resulted in significant changes in minimal inhibitory concentrations in 34% (155 of 456) of assays. These variations in activities appeared to depend more on the organism tested than on the particular aminoglycoside used; with strains of Staphylococcus aureus, Proteus, and Providenica being most affected. For all five aminoglycosides, minimal inhibitory concentrations determined by broth dilution, regardless of medium, showed poor correlation with zone sizes obtained by the Bauer-Kirby technic. These results suggest that unless some standard assay procedure for activity of aminoglycosides is adopted, meaningful comparison of results within and among laboratories will not be possible.