The paper presents the results of the statistical processing that comprises the calculation of the basic numerical statistics, the analysis of an association of signs, the estimation of Pearson's paired correlation coefficient, and the testing of its statistical significance, the construction of logit models, and the regression analysis of 350 case histories of inpatients with nephrotuberculosis treated at Samara Regional Tuberculosis Hospital One in 2002-2006. The results of the study proved it possible to simulate a clinical diagnosis for urinary tract tuberculosis. It can be simulated using the following indices: the presence of chronic renal failure, urinary leukocytes, blood urea, excretory urographic data, the electrocardiographic signs of essential hypertension, and the presence of non-tuberculous comorbidity (type 2 diabetes mellitus, gastric peptic ulcer, chronic hepatitis).