There is an urgent need for the development of standard reference methods for determining the susceptibility to antimicrobial agents of bacteria associated with fish diseases. This paper addresses the validation issues raised by the formulation of standard methods to perform disc diffusion, agar dilution and broth dilution tests designed to quantify the susceptibility of isolated bacteria to antimicrobial agents. The validation process is considered to include both internal and external components. The internal component refers to the validity of the numerical values of susceptibility produced by the methods. It is argued that robustness, intra- and inter-laboratory precision are the most important parameters of internal validity. The external component refers to the validity with which meaning can be attributed to the numerical values. The meaning of susceptibility tests is determined by the use of breakpoint values which allow the attribution of the terms sensitive or resistant to the bacteria examined. External validity, in a clinical context therefore, depends critically on the validity of any breakpoint value adopted. The theoretical and practical problems associated with setting and validating breakpoints are addressed. (C) 2001 Elsevier Science B.V. All rights reserved.