Although clinical reasoning may be very complex, it is possible to evaluate the effects of analytical variability and error for an analyte in situations in which clinical decisions are mainly based on the results of quantitative measurements of the specific analytical component. The consequences, or the validating of these consequences, can then define the quality goals for the component in the clinical situation. Two situations are investigated: a bimodal situation, ie, a situation in which the component is used for classifying patients into diagnostic groups and the diagnosis can be confirmed; and a unimodal situation, in which the risk of later development of a certain disease increases with the concentration of the component. The bimodal situation is illustrated by the application of creatine kinase isoenzyme measurements in the early diagnosis of acute myocardial infarction and by measurement of blood thyroid-stimulating hormone in screening for congenital hypothyroidism. The unimodal situation is evaluated by serum cholesterol measurement results for identifying individuals at risk regarding ischemic coronary heart disease. It is shown that this system is extremely sensitive to the effects of analytical bias and imprecision. The evaluation of clinical consequences as caused by analytical bias and imprecision makes it possible to define goals for stable analytical performance and for the maximal tolerable unstable performance when the clinical application of the test is well known.