Studies have been conducted on eight sets of monozygous and nine sets of dizygous female Negro twins, both members of whom were heterozygous for G-6-PD deficiency. Twins were studied both by assay of erythrocytic G-6-PD activity and by the methemoglobin elution test (MET). The M E T is a procedure which identifies histochemieally cells with appreciable G-6-PD activity and permits accurate determination of the percentage of such cells in heterozygotes. Monozygous twins showed significantly less "within-pair" variation than dizygous twins with both the M E T and G-6-PD assay. Concerning the significantly greater agreement in M E T results in monozygous twins than dizygous twins, our present working hypothesis is that X-chromosomal inactivation in the Negro female is genetically controlled, rather than random. However, certain alternate hypotheses allowing for random X-inactivation have not been excluded; these include somatic cell selection after random X-inactivation, and cell exchange between identical twins in utero. Studies in nontwin related heterozygotes now underway should help differentiate among these various possibilities. In addition to the studies on 17pairs of female twins heterozygous for G-6-PD deficiency, 26pairs of nondefieient female Negro twins have been studied by G-6-PD assay. Within-pair variation in monozygous twins was significantly less than within-pair variation in dizygous twins in all cases. The genetic influences detected with the G-6-PD assay in the female twins eouM theoretically be due to nonrandom X-inactivation, to genetically determined quantitative differences in enzyme activity (e.g., isoalleles),