The subsets of T lymphocytes in two infants affected with Omenn's syndrome and in 109 healthy members of the highly inbred pedigree were studied. Eighteen homozygotes in this pedigree had previously died from infection at less than 6 months of age. Both infants displayed normal numbers of peripheral blood T (E-rosette) lynphocytes, poor mitogen reactivity of lymphocytes, normal mixed lymphocyte culture reactivity, a paucity of circulating B cells, variable hypogammaglobulinemia, and elevated serum IgE concentrations. At 4 months of age, one infant (boy) had 95% T3+ (total T) peripheral blood lymphocytes, 41% T4+ (helper T) lymphocytes, and 64% T8+ (suppressor T) lymphocytes; at 4 months of age the other infant (girl) had 43% T3+, 43% T4+, 19% T8+ lymphocytes, and 18% T6+ (stage II thymocyte) lymphocytes. Age-matched controls had values of 49% for T3+, 37% for T4+, 13% for T8+, and <1%T6+. Functional lymphocyte suppression, assayed by the reverse hemolytic plaque assay, indicated that the infant girl's lymphocytes suppressed 75% of the immunoglobulin production by normal lymphocytes. One hundred and nine healthy family members (containing many obligate heterozygotes) were analyzed for distributions of T-lymphocyte subsets and compared with 37 age-matched controls. When distribution patterns of the percentage of T3+, T4+, and T8+ cells of the family members were compared to those of controls, T3+ values were lower, T4+ values were lower, and T8+ values were higher, and all differences were statistically significant by chi-square analysis ( P < 0.0005). The mean T4:T8 ratio for the controls was 2.4 ± 1 (±2 SD) and 28 family members had values below the 95% confidence level. There was general, although not absolute, correlation between a lower T4:T8 ratio in parents and appearance of the lethal trait in the offspring. Analysis of T-cell subsets in 21 obligate heterozygotes revealed that the decrease in T4:T8 ratio (1.6 vs. 2.4 for controls, P < 0.0001) was due to both a decrease in the T4 population (45 vs 53% for controls, P = 0.0038) and an increase in the T8 population (29 vs 22% for controls, P < 0.0001). It was concluded that the immunodeficiency in Omenn's syndrome is due to deregulation of T-lymphocyte subsets, appearance of immature T cells in the peripheral blood, functional T-cell suppression of immunoglobulin production, and reduced B-cell populations. An abnormal distribution of the percentage of T4- and T8-positive cells, which exists in an extraordinary number of family members, may serve as a phenotypic lymphocyte marker and, thus, aid in the identification of the heterozygous state.