A single laboratory study was carried out to compare E Test with broth microdilution and disk diffusion to establish tentative quality control ranges for Nocardia asteroides ATCC 19247 and Rhodococcus equi ATCC 6939 against a panel of eight antimicrobial agents. Reproducibility testing was performed on 12 consecutive days to establish tentative quality control ranges. A total of 36 clinical strains of the Nocardia asteroides complex and 5 Rhodococcus strains were used in the study. Both candidate control strains and clinical strains grew well on cation-adjusted Mueller-Hinton agar. Adequate growth occurred at 48 to 72 h for the Nocardia isolates and 24 to 48 h for Rhodococcus. A standardized primary inoculum of 5 x 10(4) CFU/mL was used for performance of E Test and disk diffusion for the Nocardia isolates. Tentative population-based error rates were calculated using current breakpoints for Enterobacteriaceae for E Test compared with disk diffusion for the 36 clinical strains of Nocardia species. Significant very major error rates were observed for imipenem (22%) and minor error rates varied from 2.7% to 50%. These methods require more extensive validation before definitive breakpoint criteria can be established.