There is disagreement over the identity of his parents.
Father:  William THOMPSON
Mother:  Margaret CHARLES
or
Father:  Rev. John THOMPSON
Mother:  Anne Butler (BRAYNE) SPOTSWOOD, d/o Richard BRAYNE
or
Father:  Col. Roger THOMPSON of Blackwell's Neck, Hanover Co., VA

1.  Jordan R. Dodd, ed.  1993.  KY, NC, TN, VA, WV Marriages:  Early to 1850.  Liahona Research, Orem, UT (Broderbund CD-229):  not found.

2.  Census Index:  Colonial America, 1607-1789 (Broderbund CD-310):

1779

Thompson, William

VA

Hanover Co.

This possibly is our subject.  There is no indication on the CD as to what kind of record this is.

3.  Census Index:  U.S. Selected Counties, 1790 (Broderbund CD-311):  William not found — VA census destroyed.

4.  Census Index:  U.S. Selected Counties, 1800 (Broderbund CD-312):  Frances not found — VA census destroyed, except for Accomack and Louisa Cos.

5.  Automated Family Pedigrees #1, pre-1850  (Broderbund CD-100).  United Ancestries #UA/MEMMOTT 1500058.

6.  K.C. Gottschalk & J.B.C. Nicklin.  1981. " The Pendleton Family."  Pp. 646-780 in Genealogies of Virginia Families: from the Virginia Magazine of History and Biography. Genealogical Publ. Co., Baltimore (Broderbund CD-162).

7.  Horace Edwin Hayden.  1891.  "Ball, of Virginia."  Pp. 45-144 in A Genealogy of the Glassell Family of Scotland and Virginia... (self-published) Wilkes-Barre, PA (Broderbund CD-205).

8.  Beverly M. Stercula.  [1960s-1970s.]  Thompson Family Magazine.  Archives of the THOMPSON-L mailing list (online at RootsWeb.com).

9.  Beveryly M. Stercula.  [1966.]  The Thompson Families of Hanover, Louisa, Albemarle, Goochland, Amelia, and Fluvanna Counties in Virginia.  (Unpubl. manuscript partially online in the Archives of the THOMPSON-L mailing list at RootsWeb.com)  Stercula shows nine children for William and Frances, but no son "John" or "Thomas" or "John Thomas."

10.  Broderbund.  World Family Trees.  Vol. 13, Pedigree No. 1281; Vol. 14, Pedigree No. 2167.

11.  LDS.  Family Search: Internet Genealogy Service: International Genealogical Index (IGI).

12.  George Rose & Margaret Rose.  1969.  The Kay-Pendleton-Neel Families.  J. Grant Stevenson, Provo, UT (online at GenealogyLibrary.com).

Frances Mills, daughter of Charles Mills and Ann (Jackson?) Mills, was born in Hanover County, and was married there on January 29, 1771, to William Thompson, Esq.  They took up residence in Saint Mark's Parish, Culpeper Co.  William Thompson, Esq., died May 3, 1797, and his wife, Frances Mills, died November 19, 1802.  They were married January 29, 1771.  (Hayden, Virginia Genealogies, 131).  Their Bible record deposited in the DAR Library, Washington, D.C., names the following children:  Peggy Thompson born February 6, 1772 Charles Thompson born March 7, 1773 William Mills Thompson born January 11, 1775 Ann Thompson born June 18, 1777 Sarah Mills Thompson born December 15, 1779 Mary Ann Thompson born December 1, 1782 Frances Jackson Thompson born December 29, 1784 Edmund Thompson born April 11, 1787 Nathaniel Thompson born August 25, 1789 ... An account of the father of Frances Jackson Thompson was written by Judge John W. Jones of Bowling Green, Kentucky, in June of 1900, which article is reprinted in Genealogical & Historical Notes on Culpeper County, Virginia (p. 86), by Ralph Travers Green.  This article shows that William Thompson, the father of Frances Jackson Thompson, came from England to the United States and settled in Hanover County, Virginia, probably in the second half of the Seventeenth Century.  William Thompson appears in Spotsylvania records as early as 1745.  He married Frances, daughter of Charles Mills, on January 29, 1771.  The William Thompson who married Frances Mills is sometimes referred to as William Thompson, Jr.  This does not necessarily prove that his father was also William Thompson because in the Colonial period the term Junior simply served to indicate a person younger than another of the same name but not necessarily a son.  William Thompson appears as a tithable in St. Martin's, Hanover County, 1793, though it seems that his children were in Orange and Culpeper Counties.  There is a typewritten manuscript in the Virginia State Library by Henry Lockhart and entitled "Thompson of Hanover and Louisa Counties" (Cs 7i--T47- 1960).  This source says that tradition regards William Thompson, Jr., as the son of Colonel Roger Thompson of Blackwell's Neck of Hanover County and that this Roger Thompson was Sheriff of New Kent County in 1706.  He shows that in 1790 William Stewart of Sussex County sold to William Thompson and his wife Frances 100 acres of land adjoining Christopher Smith, Robert Morris, and William Thompson.

How can our subject have appeared in Spotsylvania Co. in 1745 when he wasn't born until 1749?!

11.  National Society, Daughters of the American Revolution.  19??.  Lineage Books of the Charter Members of the National Society of the Daughters of the American Revolution.  Vol. 39.  NSDAR, Washington, DC (online at Ancestry.com), p. 353.

12.  WorldConnect (online at RootsWeb.com).

The pedigree found on the United Ancestries CD (see Sources) has been widely circulated, which is unfortunate because it cannot possibly be correct, at least with regards to the son "John Thomas."  First of all, I can find no record of any son named "John" or "Thomas" or "John Thomas" for William and Frances (MILLS) THOMPSON, but even if they had such a son, he could not possibly be the same man as the Thomas THOMPSON who was born 7 Jul 1797 in KY and who married Delilah DAVIS because: 

1.  The father of Thomas THOMPSON, husband of Delilah DAVIS, has been proven to be Peter THOMPSON. 

2.  The family bible of William and Frances (MILLS) THOMPSON lists nine children, none of whom were named John, Thomas, or John Thomas and none of whom were born after 1789.  This fact, alone, is enough to prove that Thomas THOMPSON (1797-1872) was not their son.

3.  Thomas THOMPSON was born 7 Jul 1797, and he had a sister, Rachel THOMPSON, who was born 17 Jul 1803.  William died in 1797 and Frances died in 1802, so they could not possibly be Rachel's parents nor, by inference, the parents of Thomas.  The fact that William died in May 1797 is what is behind the statement sometimes made that "Thomas THOMPSON's father died before he was born."  Not so.  Thomas THOMPSON's actual father, Peter THOMPSON, died in Missouri in the 1830s. 

4.  William THOMPSON was born in and died in Virginia.  To the best of my knowledge, there is no evidence that either William or Frances was ever in Kentucky.  Both Thomas and Rachel were born in Kentucky.  Through no less than three censuses (1850/60/70) and his 1867 voter registration, Thomas consistently gives his birthplace as Kentucky, as does his biographer, E.B. Ware (1916).

5.  If Frances (MILLS) THOMPSON were Thomas's mother, she would have been 47 or 48 years of age at his birth and there would have been an eight-year gap between her last known child (Nathaniel) and "John Thomas" — not impossible, but both her age and the gap make her an unlikely candidate for Thomas's mother.    And even if the death dates for both William and Frances are wrong and they were both still alive in 1803, I simply can't buy Frances as birthing Rachel at age 53 or 54.

6.  To the best of my knowledge, the Thomas THOMPSON who was born 7 Jul 1797 in KY was never in his lifetime called "John Thomas." Only in recent decades have people referred to him by that name (i.e., as far as I know, no source contemporary to Thomas called him "John Thomas").  If anyone has a contemporary source calling him "John Thomas," I would very much appreciate being made aware of it.

I'm told that the roots of this error are to be found, decades ago, in the efforts of a descendent of Thomas THOMPSON to join the D.A.R. and that, while this membership was originally accepted, the D.A.R. later rejected it due to the lack of proof of the connection between Thomas and these parents.  The great irony of this episode is that descendants of Thomas and Delilah (DAVIS) THOMPSON would be easily eligible to join the D.A.R. through Delilah's maternal grandfather, Col. Henry CLARK.