Klinefelter (47,XXY) syndrome occurs in approximately 1:800 male births and accounts for about 10-20% of males attending infertility clinics. Recent studies have shown no obvious phenotypic differences between subjects in which the extra X-chromosome is of paternal or maternal origin; however, a minority of Klinefelter patients are adversely affected clinically and intellectually to an exceptional level, and the underlying basis of this phenotypic variation is not known. We hypothesize that skewed X-inactivation and possibly parental origin of the X-chromosomes is involved. In this study, we determined parental origin and inactivation status of the X-chromosomes in 17 cytogenetically confirmed 47,XXY cases, two 48,XXYY cases and one mosaic 46,XY/47,XXY case. Eight highly polymorphic markers specific to the X-chromosome and the polymorphic human androgen-receptor (HUMARA) methylation assay were used to determine the parental origin and X-inactivation status of the X-chromosomes, respectively. Overall, 17 cases were fully informative, enabling parental origin to be assigned. In 59% of cases, both X-chromosomes were of maternal origin (Xm); in the remaining 41%, one X was of maternal (Xm) and one was of paternal origin (Xp). In 5 of 16 (31%) cases informative at the HUMARA locus, skewed X-inactivation was observed as defined by greater than 80% preferential inactivation involving one of the two X-chromosomes. The two 48,XmXpYY cases both showed preferential paternal X-chromosome (Xp) inactivation. Three 47,XmXmY cases also showed preferential inactivation in one of the two maternal X-chromosomes. These results suggest that skewed X-inactivation in Klinefelter (47,XXY and 48,XXYY) patients may be common and could explain the wide range of mental deficiency and phenotypic abnormalities observed in this disorder. Further studies are warranted to examine the role of X-inactivation and genetic imprinting in Klinefelter patients.