By using agar dilution as the standard method, we determined the ability of broth microdilution, disk diffusion, and an automated system (AMS Vitek) to detect three different levels of vancomycin resistance among six enterococcal isolates. Enterococcus gallinarum AIB-38 and AIB-39 exhibited low-level resistance (MIC, 16 to 32 micrograms/ml) that was detected only by agar and broth dilution methods. E. faecalis V583, E. faecium AIB-42, and E. faecalis AIB-41 showed moderately high-level resistance (MIC, 128 to 256 micrograms/ml) that was detected by dilution methods but not by disk diffusion unless prolonged incubation or an inoculum 10-fold greater than the standard was used. Similarly, this level of resistance was detected by AMS Vitek only when an inoculum 10-fold larger than recommended was used. The high resistance demonstrated by E. faecium AIB-40 (MIC, 2,048 micrograms/ml) was readily detected by all methods studied. Variation in the ability of different methods to detect vancomycin resistance among enterococci complicates monitoring the incidence of these organisms and could result in very major susceptibility reporting errors.