We have identified 21 different alpha-galactosidase A gene (GLA) mutations in 22 unrelated Czech and Slovak families with Fabry disease. Eleven of these mutations were novel (point mutations D93N, A135V, D155H, G171R, Q280K, G360S, Q330X, splicing errors c.194ins14, c.801ins36 and deletions c.674_732del59, g.3405_6021del2617). Genotyping of family members for family-specific mutations revealed 55 heterozygotes that manifested clinical symptoms of different severity. To examine the contribution of X-inactivation skewing to disease manifestation in Fabry heterozygotes, we have adopted the Mainz severity scoring scheme and compared the score values with the X-inactivation status in 39 carriers in an age-dependent manner. The age-score trendline of Fabry females who had a predominantly inactivated X-chromosome bearing a wild-type GLA allele (10 of 38 females) was markedly steeper than in the rest of the cohort. One female carrier with an inactivated mutated allele had a low score value when compared to the other heterozygotes of the same age. These data suggest that X-inactivation is indeed a major factor determining the severity of clinical involvement in Fabry heterozygotes. There was a statistically significant difference between the severity score values of heterozygotes with random and non-random X-chromosome inactivation at the 5% level of significance. Further studies will show if the degree of the wildtype allele inactivation will be useful as a predictive marker of severity of phenotype in Fabry heterozygotes. Although the correlation between X-inactivation skewing and presentation of the disease in Fabry heterozygotes has previously been suggested in the literature, this report is among the first attempts to examine this relationship systematically.