The informational efficacy is evaluated of diagnostic tests that are employed in both screening programs and clinical routine. The main data concerning with the decision making theory are summarized and may be successfully applied in order to obtain diagnoses in practical conditions. Various diagnostic algorithms are demonstrated on practical examples. The attention is focused on such diagnostic methods which are issued from binary input data, ROC curve, the use of information theory and Bayes formula. Also the problem related to the statistical induction in evaluating diagnostic tests is treated.