Jefferson's political enemies, and some biographers, have claimed that Jefferson had a sexual relationship with a slave named Sally Hemming, and that the two produced children.
There is a fair chance that the issue can be settled with DNA testing. Word is it has been, that the results have been submitted to a journal.
The method of proof would be to compare claimed black decendents with known white decendents. However, with the two branches of the family now separated from one another by seven or eight generations at least (four generations counting from the present back to Thomas along one line, then four more generations back down the other line to presently living people), the problem presents special difficulties.
In principle one could compare pedigrees using the same technique illustrated for a complicated Okinawan family namely, analysis with the kinship program. However, across such a great distance of generations information from standard (autosomal) DNA markers would be so diluted it would be hard to sort from noise. It could be done either by testing hundreds of RFLP markers (preferably by sequencing), or by testing scores of people on each family branch, or by a combination. But still it would be difficult.
Fortunately, there is an ingenious alternative. Jefferson's Y-chromosome would not be diluted by the passage of generations. Rather, to each offspring it is either passed intact (if the offspring is a boy), or not passed at all (to a girl). Hence, if Jefferson had a son from his white wife, who had a son, who had a son ... who had a son alive today, then from that patrilineal decendent the Y-chromosome traits are available just as from Jefferson himself. Then, assuming that among the decendents of Sally Hemming who may be alive today and may be decended from Jefferson, there are some who are patrilineal, they should also have the same Jeffersonian traits at loci on the Y-chromosome.
Sound too easy? Well, there is at least one hitch. One alternative to the theory that Jefferson fathered Sally Hemming's children, is that his nephew did. If the nephew in question was Jefferson's brother's child, then he too would share the infamous Jeffersonian Y-chromosome, and Y-locus tests therefore would not distinguish uncle from nephew as possible progenitor.
We anxiously await publication of the completed research.
The paper reporting the results is under submission, still not published. It appears that at least two lineages are examined, and Jefferson is the father in one but not the other.
After rejection by two prestigious journals (too hot to handle?), the paper was finally accepted by Nature, the most prestigious journal of all.