Desirable standards of performance of laboratory tests, termed analytical goals, are required for use in quality assurance, evaluation of methods, reagent kit sets and instruments, discussions with clinicians and gaining additional laboratory resources. Traditionally, goals have been derived from fractions of the reference interval, opinions of clinicians, the state of the art, views of individuals and groups and data on biological variation. All have disadvantages, but the last is currently favoured by many as the best strategy to delineate general goals. Recent more novel approaches have been concerned with the definition of goals for particular clinical situations, but these have not been widely accepted as yet. Further work is required on the setting of goals for performance characteristics other than imprecision and inaccuracy and on the use of goals in the design of effective quality control procedures.