Antibiotic sensitivity testing has been developed into a highly sophisticated process technically. Interpretation of results obtained requires the definition of breakpoints in order to separate sensitive (likely to respond to treatment) from resistant (likely to persist despite treatment) strains. Various methods used for the setting of breakpoints are described and their merits and drawbacks discussed. Although there may be wide disparities in the definitions adopted in different countries, there is no clinical evidence that favours one set of figures over another. Problems over determination of breakpoints for fungi and of antibiotic combinations remain to be resolved.