Regrowth of aminoglycoside-resistant variants was seen when large inocula of two strains of Escherichia coli were incubated with gentamicin in concentrations well above their MICs (0.5 ,ig/ml). The extent of the selection of resistant variants was proportional to the concentration of gentamicin during incubation; after incubation with gentamicin (22 ,ug/ml for 24 h), all bacteria were resistant to at least 8 ,ug/ml. Bacteria resistant to these concentrations always formed small colonies, whereas variants resistant to lower concentrations (1 to 2 ,ug/ml) could form both small and normal colonies. The regrowth of resistant variants could be monitored by luciferase assay of intracellular ATP in cultures incubated with gentamicin (c2 ,ug/ml). In cultures incubated with higher concentrations, regrowth did occur, although this did not result in viability (CFU per milliliter) or ATP levels above those of the initial inocula. The implications of this regrowth for MIC determinations in broth and the possible clinical revelance of the resistant variants are discussed.The endpoint in MIC determinations, i.e., visible growth after overnight incubation, is the net result of a series of possible events occurring during a long period of incubation (9).These events include the initial killing of bacteria, liberation of enzymes that degrade the antibiotic, pH and other media changes that affect the activity of the antibiotic, and finally the possible regrowth of bacteria that have survived the exposure to the antibiotic. All of these factors may differ among different bacteria and antibiotics and are in general difficult to control.Regrowth of bacteria is seen quite frequently when gramnegative enteric bacteria are exposed to aminoglycosides (5). This regrowth seems to be due to the selective multiplication of resistant variants that are present in a low frequency in the initial inoculum (5,14,25). Such a regrowth obviously complicates the interpretation of MIC determinations performed in broth. The phenomenon could account for some of the following discrepancies observed in MIC determinations against aminoglycosides: (i) different MICs obtained by rapid methods and conventional methods with overnight incubation, (ii) different MICs obtained in broth and solid media (which may also be due to differences in cationic content [26]), and (iii) the pronounced effect of inoculum size on the MIC.The demonstration of regrowth by viable count is cumbersome. The possibility of monitoring the regrowth by the firefly bioluminescence assay of intracellular ATP (24) was therefore explored. Intracellular bacterial ATP increases concomitantly with viability (CFU per milliliter) (18,19), and the assay of intracellular ATP is therefore a sensitive method for monitoring bacterial growth.In this work the quantitative aspects of regrowth in aminoglycoside-exposed cultures was studied by determination of viability combined with population analysis and also by assay of intracellular ATP. (ii) Disk diffusion. Disk diffusion assays were performed with 3...