Epidemiologic studies of rare diseases may produce surprising findings and raise ethical issues. This is illustrated in this study performed in 171,977 consecutive Taiwanese newborns (including 90,288 boys) from July 2006 through June 2008 by measuring dry blood spot and then leukocyte α-galactosidase A (α-Gal A) activities and finally by detecting mutations in the GLA gene involved in Fabry disease. Schematically, two phenotypes of this disease are known: the classic form, with systemic involvement and very low α-Gal activity in males; and the later-onset form (>40 years of age), with some residual α-Gal A activity, which is dominated by cardiac involvement. All 11 newborns who had <5% of normal mean α-Gal A activity were boys who had GLA mutations. In the group of 66 newborns (64 boys and 2 girls) with α-Gal A activities between 5% and 30%, 61 hemizygous boys and 2 heterozygous girls had GLA mutations. Among the group of 12 newborns (11 boys and 1 girl) with α-Gal A >30%, only 1 boy had a previously reported mutation, identified in a family with later-onset renal disease. In total, 72 male and 2 female newborns had GLA mutations, an overall frequency of approximately 1 in 1250 boys and approximately 1 in 40,840 girls. Four boys were “predicted” to have the classic phenotype, a frequency of about 1 in 22,570 newborn boys. In contrast, the estimated frequency of the later-onset phenotype is approximately 1 in 1390 male newborns mutations. Three families provided information on other members of the kindreds, two with classic and one with later-onset phenotype. All three families had previously undiagnosed symptomatic family members, including one heterozygous female with ESRD and two males with renal involvement. This is undoubtedly the positive side of such studies. A second study was performed more recently in 110,027 Taiwanese newborns between January 2008 and January 2009 by using a similar protocol (plasma α-Gal A activity was measured) (1). The results of this study confirmed those of the previous screening. A high prevalence of the cardiac variant Fabry mutation IVS4 + 919G→A, first discovered in Japanese patients, was found among newborns (approximately 1 in 1600 boys) in both studies. This splicing mutation was most common (82% of patients). The alternatively spliced transcript was normally present in small amount (<5% of normal transcript) in most human tissues. However, the G→A transversion enhanced the percent expression of the alternatively spliced α-Gal A variant and included a 57-nucleotide intronic sequence that caused a frameshift mutation, resulting in a truncated enzyme polypeptide that had no detectable enzyme activity. The clinical significance of this splicing mutation remains to be fully clarified. Of interest, Lin et al. (1) have investigated 9 grandfathers and 11 grandmothers carrying this mutation, as do their respective grandsons. Among the 9 maternal grandfathers, only 3 had hypertrophic cardiomyopathy, compared with none of the 11 grandmothers. These results should be compared with those reported in 2006 from Torino, Piedmont, Italy, by Spada et al. (2). They screened 37,000 consecutive newborns with similar methods and identified 12 infants with GLA mutations, including 11 who had molecular lesions that expressed residual activity consistent with the later-onset phenotype. The overall frequency of Fabry mutations was approximately 1 in 3100 Caucasian boys. Mutation analysis predicted that one of the newborns had the classic phenotype (1 in approximately 37,000), whereas 11 of the newborns were predicted to have the later-onset phenotype (1 in approximately 3400). The prevalence of the classic form is close to that found in previous estimations, whereas that of the later-onset phenotype seems to be higher than commonly thought. In Taiwan, the frequency was 2.5 times more frequent than in the Italian population. These studies raise many ethical and clinical issues. What can be said to the parents of neonates harboring a GLA mutation suggestive of a later-onset disease? The clinical consequences of some mutations, if any, cannot be predicted. The ethnic background should be taken into account. If clinical consequences can be expected, when to start evaluating the cardiac condition and when to consider enzyme replacement therapy, if necessary? What are the psychologic consequences of the screening on the mutation carrier and his/her family? The ethical issues raised by early detection and prediction of later-onset genetic disease should be debated, not only by experts, but also with the general population.