The E Test (PDM Epsilometer; AB Biodisk, Solna, Sweden) was compared with microbroth dilution and disk diffusion for antimicrobial susceptibility testing of Nocardia as a collaborative study by two geographically separate laboratories. A total of 52 clinical isolates and five species of Nocardia were used in this comparative evaluation. Susceptibility testing was performed with Mueller-Hinton media and eight antimicrobial agents. Growth of the test strains with Mueller-Hinton medium was generally satisfactory, with the majority of isolates producing adequate growth within 24-36 h. Growth inhibition ellipses were generally well delineated and uniform for most drugs, and the points of intersection with the E Test strip were generally easy to determine. An inoculum size of approximately 2.0 x 10(7) CFU/ml was optimal for performance of the E Test method with the Nocardiae. Comparison of E Test and microbroth dilution MICs revealed 89.4% agreement for all drugs within +/- 1 log2 dilution. Using NCCLS interpretive criteria for susceptible and resistant results, complete agreement between E Test and disk diffusion results was 93.3%, and between E Test and microbroth dilution results was 96.2%. Interpretive category errors occurred at rates of 18.2% (risk corrected), 0, and 4.1% for very major, major, and minor errors, respectively, when E Test results were compared with disk diffusion results, and 0, 0, and 3.8%, respectively, when E Test was compared with microbroth dilution. Inter- and intra-laboratory reproducibility, within +/- 1 log2 dilution for all drugs, was 95% and 98%, respectively. Results from this study suggest that E Test may be suited for use as an alternative method for antimicrobial susceptibility testing of Nocardia species.