Nephrogenic diabetes insipidus (NDI; designated 304800 in Mendelian Inheritance in Man) is an X-linked disorder with abnormal renal and extrarenal V2 vasopressin receptor responses. The mutant gene has been mapped to Xq28 by analysis of RFLPs, and tight linkage between DXS52 and NDI has been reported. In 1969, Bode and Crawford proposed, under the term "the Hopewell hypothesis," that most cases in North America could be traced to descendants of Ulster Scots who arrived in Nova Scotia in 1761 on the ship Hopewell. They also suggested a link between this family and a large Mormon pedigree. DNA samples obtained from 13 independent affected families, including 42 members of the Hopewell and Mormon pedigrees, were analyzed with probes in the Xq28 region. Genealogical reconstructions were performed. Linkage between NDI and DXS304 (probe U6:2.spl), DXS305 (St35-691), DXS52 (St14-1), DXS15 (DX13), and F8C (F814) showed no recombination in 12 families, with a maximum lod score of 13.5 for DXS52. A recombinant between NDI and DXS304, DXS305, was identified in one family. The haplotype segregating with the disease in the Hopewell pedigree was not shared by other North American families. PCR analysis of the St14 VNTR allowed the distinction of two alleles that were not distinguishable by Southern analysis. Carrier status was predicted in 24 of 26 at-risk females. The Hopewell hypothesis cannot explain the origin of NDI in many of the North American families, since they have no apparent relationship with the Hopewell early settlers, either by haplotype or by genealogical analysis. We confirm the locus homogeneity of the disease by linkage analysis in ethnically diverse families. PCR analysis of the DXS52 VNTR in NDI families is very useful for carrier testing and presymptomatic diagnosis, which can prevent the first manifestations of dehydration.