The BSAC Working Party on Susceptibility Testing has developed a standardized method of disc susceptibility testing that has been ‘field tested’ in 19 diagnostic laboratories in the UK and Ireland. The method employs semi-defined media, a semi-confluent inoculum and relates zones of inhibition with BSAC-specified MIC breakpoints to interpret susceptibility. The bacteria selected for the trial included clinical isolates and control strains from the ATCC and NCTC national collections. Organisms were chosen because they had known attributes, such as being fully susceptible or having a demonstrated mechanism of resistance. The results from this survey are very encouraging. With the commonly isolated Enterobacteriaceae, specifically Escherichia coli, Proteus mirabilis and Klebsiella spp., no major problems were observed except with gentamicin and cefuroxime. In the case of gentamicin, problems were associated with resistant strains of P. mirabilis, with MICs of 2 mg/L, being falsely reported as susceptible. For cefuroxime, it is not unexpected that results were unreliable, as the MIC distribution straddles the in vitro breakpoint concentration (following the results of this study the MIC and zone diameter breakpoints have been amended to improve reporting). No major problems were encountered for Pseudomonas aeruginosa with the agents studied. The ‘field survey’ has shown that disc testing is unreliable for determining the susceptibility of coagulase-negative staphylococci to teicoplanin, and that the detection of glycopeptide resistance in enterococci is improved by incubation for a full 24 h. Inconsistencies observed with fastidious organisms were associated with incorrect inocula. Zone diameter data for the control strains studied provide information that can be utilized by diagnostic laboratories to monitor the daily performance of testing.