Methods for detection of high-level resistance to aminoglycoside-aminocyclitol antibiotics were evaluated using 104 blood isolates of enterococci (97 Enterococcus faecalis and 7 Enterococcus faecium). Kanamycin was used to predict resistance to amikacin. Discrepancies between methods were resolved by time-kill studies. Four methods (MicroScan, macrotube, microtiter, and disk diffusion) for detecting resistance to gentamicin and streptomycin were compared, using 51 consecutive strains. There were 13 gentamicin-resistant strains, all of which were detected by macrotube, microtiter, and disk diffusion. MicroScan detected 2 (15%) of the 13. Of the 18 streptomycin-resistant strains, 17 (93%) were detected by disk diffusion, 16 (89%) by microtiter, 9 (50%) by macrotube, and 6 (33%) by MicroScan. An additional 53 consecutive strains were examined only by disk diffusion and microtiter for resistance to gentamicin, streptomycin, and kanamycin. The entire population of 104 strains contained 35 gentamicin-, 22 streptomycin-, and 54 kanamycin-resistant enterococcal isolates. All 35 gentamicin-resistant strains were detected by both methods. Of the 22 streptomycin-resistant strains, 1 was detected only by microtiter, 2 only by disk diffusion, and 19 by both methods. Of the 54 kanamycin-resistant strains, 1 was detected only by microtiter, 2 only by disk diffusion, and 51 by both methods. One additional strain which was resistant only by disk diffusion was susceptible to amikacin plus penicillin by time-kill studies. Disk diffusion is a suitable method for detection of high-level aminoglycoside-aminocyclitol resistance in E. faecalis and is well suited for sporadic testing. Additional data are necessary to determine the suitability of these tests for E. faecium.