The electrophoretic variants of G6PD have provided evidence for a single active X-chromosome in human somatic cells, but two active X-chromosomes in germ cells and in triploid cells with a 69,XXY karyotype. Studies of clonal populations of cells heterozygous for G6PD-A but expressing a single isozyme have provided evidence that the maintenance of the inactivation process is extremely stable, but that an occasional localized derepression event may occur. Such populations have also been used to show that methylation of X-chromosome DNA is not obviously different for XA and XI; furthermore, methylation is surprisingly unstable at least at some sites on the X. Studies of derepressed loci on the inactive X-chromosome in these clonal cell populations show that the expression of loci that escape inactivation is greater on XA than XI. G6PD variants have provided significant insights into the consequences of X-chromosome inactivation, revealing the role of intercellular communication and cell selection in determining female phenotype. The evidence that is now available indicates that the regulation of the X-chromosome is complex. The mechanisms involved include: 1) initiation, an event that may be mediated through an autosomal product and 2) maintenance which is stable but subject to programmed derepression of the entire chromosome during ontogeny of germ cells and occasional unprogrammed localized derepression in somatic cells. In addition, there may be transcriptional differences between XI and XA to compensate for monosomy of X-linked genes. The molecular basis for this multilevel regulation is unclear, but it seems certain that the cloned DNA probe for G6PD [Persico et al, 1981] will provide further insights.