Screening agars containing different vancomycin concentrations and different susceptibility testing procedures were compared to determine their validity for the detection of glycopeptide-resistant enterococci (GRE). Direct streaking of rectal swabs over the surface of a commercially available agar (Enterococcosel; Becton Dickinson, Germany) containing 4 microg/ml and 16 microg/ml vancomycin was followed by incubation for 24 to 48 h. Susceptibility tests were done by microbroth dilution, disk diffusion, and the E test (AB Biodisk, Sweden). The microbroth dilution method according to National Committee for Clinical Laboratory Standards (NCCLS) was used as the gold standard for detection of GRE. Resistant and intermediately susceptible enterococcal isolates were differentiated to the species level. To detect resistance genes, the polymerase chain reaction was performed on all intermediately resistant isolates, on all isolates of VanB phenotype, and on 30% of isolates of VanA phenotype. Screening agar containing 4 microg/ml vancomycin displayed high sensitivity (97.6%) but only low specificity (35%) for the detection of GRE. Screening agar containing 16 microg/ml vancomycin had a high specificity of 89.3% and only a slightly lower sensitivity (92.7%) than the screening agar containing 4 microg/ml vancomycin. For the disk diffusion test, a breakpoint of < 16 mm yielded the optimal combination of sensitivity (98.8%) and specificity (99.6%). Both sensitivity and specificity of the E test for GRE detection were 100%. However, the E test is too expensive for testing of all enterococci. In conclusion, the combination of an inexpensive screening agar with either the E test or the disk diffusion test constitutes a valid and cost-effective method for the detection of GRE from screening specimens.