A series of experiments were performed to investigate the impact of the inclusion of internal controls on the inter-operator precision of disc diffusion assays. The susceptibilities of 98 strains of Aeromonas salmonicida to florfenicol were determined by ten untrained operators using two disc diffusion test protocols. One was the Clinical Laboratory and Standard Institute's M42-A protocol which, in its quality control requirements, employs external controls. The second was a version of the protocol published by Alderman and Smith in 2001 that had been modified to include internal controls. The only alteration of the protocol required by this modification was the inoculation of the control strain onto one half of an assay plate and the test strain on the other and the placing of the discs between the areas inoculated with the two strains. The output data of the modified, internal control, method was the difference in the radii of the control and test strains. There was significant inter-operator variation in the data generated using the M42-A protocols. For replicate (n=9-11) measurements of the control strain, the mean of the values generated by the untrained operators ranged from 33.5 mm-45.1 mm. The mean zone sizes for the 9-11 wild-type test strains each operator tested varied from 35.0 mm-45.2 mm. The extent of inter-operator variation was significantly reduced when the protocol incorporating internal controls was used. Using this modified protocol, the range of the operator means for the difference in radii for wild-type strains was 0.2 mm to 1.4 mm.It is argued that the modification of a standard protocol to include internal controls would allow the classification of a clinical isolate using a single plate and that this would have advantages for laboratories handling a limited number of strains a year. This initial investigation of the performance characteristics of such a modified protocol suggests that more extensive studies are warranted. (C) 2008 Elsevier B.V. All rights reserved.