Heterozygosity at nine genetic loci (PI, TF, PGM1, ACP1, HP, GC, GLO1, C3, and ESD) was analyzed in pulmonary tuberculosis patients with good (group 1, N = 71) and poor (group 2, N = 35) response to treatment. The observed heterozygosities were compared with the expected values, which were calculated from allele frequencies in a control sample of healthy individuals (N = 328 with all but one locus and 78 with ESD) according to Hardy-Weinberg expectations. The analysis showed that the observed heterozygosities gl of patients significantly differed from the expected values hl in the case of four loci (GC, PI, C3, and ACP1). The observed heterozygosity was higher than expected in three cases (PI, C3, and ACP1) and lower then expected (GC) in one case. When data on each individual locus were compared using Fisher's exact test, both groups of patients proved to significantly differ (PF < 0.05) from the control group in the same four loci. No difference in observed heterozygosity was detected between the two groups of patients. The mean expected heterozygosity was h = 0.386 +/- 0.00674; the mean observed heterozygosity was g = 0.415 +/- 0.02 in group 1, g = 0.402 +/- 0.026 in group 2, and g = 0.371 +/- 0.00955 in the control group. The t test did not reveal a significant difference between the mean values of expected observed heterozygosities. Heterozygosity at individual loci, rather than mean heterozygosity, was proposed as an integral nonspecific indicator of the genetic control of a disease, because the former directly implicates individual marker loci in the development of a disorder, whereas effects of individual loci may eliminate each other when mean heterozygosity is computed. Based on the results obtained, a genetic control was assumed for the development of the tuberculosis process in the lungs.