Abstract
A deoxyribonucleic acid marker linked to the locus for Huntington's chorea exists, but its possible use in the prediction of this disorder depends on the pedigree structure of individual families. Analysis of data from a population register for Huntington's chorea in south Wales showed that only a minority of subjects at risk had the appropriate members of their family living to allow the presence or absence of the gene to be definitively predicted. However, the structure of the family allowed a degree of prediction (in particular, exclusion of the disorder) to be made for the fetus during pregnancies of these subjects in almost 90% of cases. Such a prediction need not alter the risk state for the parent at risk. The structure of the family will remain crucial for prediction even when current limitations of the linked marker have been overcome.