Prenatal diagnosis was requested by an obligate carrier of a new syndrome of X-linked mental retardation. There was close linkage between the disease gene and the hypervariable VNTR marker DXS255 with a lod score of 4.82 at O=0 (90% support interval 0.00-0.12). When the request for prenatal diagnosis was made, additional family members were examined, resulting in an amended lod score of 6.71 at O=0.0 (90% support interval 0.00-0.09). There were no informative flanking markers at the time of the request for prenatal diagnosis; hence it proceeded by 2 point linkage analysis. The fetus was female with a carrier risk in the interval of 91-100%. Given the limitations of the mapping data available for this disorder at the time of the request, the options of accepting or rejecting this as a case for prenatal diagnosis were carefully considered. Whilst prenatal diagnosis based on fetal sexing would be sufficient to prevent the birth of an affected child, the magnitude of the known two-point lod score between DXS255 and the disease gene provided a means for diagnosis with an accuracy between 91 and 100%.