A rational framework is presented for establishing valid breakpoint concentrations to be employed in determining the frequency of oxolinic acid resistant bacteria in marine environments. There is a shortage of relevant data concerning the sensitivities of truly marine fish pathogens and, therefore, quantitative estimates of appropriate breakpoints have been made by reference to data on the sensitivity of Aeromonas salmonicida. The use of these data may limit the general applicability of the breakpoints established. Consideration of clinical, pharmacokinetic and microbiological data indicate that the incorporation of 0.5 mu g/ml oxolinic acid in Tryptone Soya Agar (TSA) would provide a suitable breakpoint for selecting resistant strains of A. salmonicida. Arguments are presented that the media currently being employed in studies of the frequency of resistance in the vicinity of marine fish farms are inappropriate. It is suggested that media specifically formulated for work with antimicrobial agents, such as Iso-Sensitest Agar (ISA) and Mueller Hinton Agar (MHA) would be more suitable. Seawater is routinely added to media used to enumerate bacteria from the marine environment. The ions present in seawater, particularly Mg2+ and Ca2+, inhibit the antimicrobial activity of oxolinic acid. When 70% artificial seawater is used in preparing ISA, the median percentage bioactivity of oxolinic acid is approximately 4.5%. Thus, in this medium oxolinic acid concentrations of 10-12 mu g/ml will exert a selection equivalent to that exerted by 0.5 mu g/ml in standard TSA media. (C) 1998 Elsevier Science B.V. All rights reserved.