This family seems to have first settle in Provence
and have been "barbarians" or aliens hence Barbee,
Presumably they were Viking companions of Rollo and later William the Conqueror.

?m2  21 Jan 1783 Hannah Barker

Barbee's Tavern - On the northwest corner of the Crossroads is a yellow stuccoed frame building (old log tavern) with a center stone chimney, which was Barbee's tavern, dating from the pre-stage coach days, circa 1787 built by Joe Barbee. The tavern was in operation during colonial and Civil War times and apparently had a thriving business because of its location. The building has wide pine flooring, exposed American Chestnut interior logs, and 3 stone fireplaces. When the tavern was converted to a house, while in the process of removing many of the walls, the owners, Mr. and Mrs. Thomas Furr (Joe Barbee's grand daughter and her husband), found that the walls were lined with old newspapers. It is believed this was also where Barbee's Post Office was. (Thanks to Lita Stickler for the history of Hume and Barbee's Tavern.) 

The Barbee DNA Project is producing information which is useful in directing further genealogy research. We have 
determined that the early Andrew and William Barbee lines are NOT related. This means that there are at least 2 distinct Barbee lines that have different DNA patterns in the U.S. and that a DNA test will show whether you are related to one of these two lines. 
identified previously unconnected Barbee lines that are related to the Andrew or William Barbee lines. This helps those families narrow their genealogy research.

Recently found genetic differences between related Barbee lines which may be helpful in narrowing down where family connections occurred.
It has been determined that the Barby line from England IS NOT related to the Andrew and William Barbee lines in the U.S. 
As of November 2008, we have received DNA results for 25 participants shown in the tables below including: 
5 known descendants of William Barbee and 14 other Barbee families that are related to William Barbee. 
2 known descendants of Andrew Barbee and 3 other Barbee families that are related to Andrew Barbee. 
1 known descendant of Edward Barby, born 1691 in England. 
 November 2008 - The most recent new participants are kits 107265 and 107308 who are part of the William Barbee line and kit 93283 who traces his Barby ancestry back to 1691 in Wicken, Northamptonshire, England. I recruited the English Barby on a hunch that he would be related to either the Andrew or William Barbee lines -- this proved to be incorrect. This is the first European Barbee (or variation) that we have tested. In this phase of the project, I would like to emphasize recruiting English and French Barbees (or variations) to prove or disprove some of the theories floating around about our origins.

Unique mutations may help identify what branch you connect to. For example, my John "Smith" Barbee line is the only part of the William Barbee line so far that has a mutation on DYS marker 454. My first cousin once removed has the same mutation as my brother. Because of the uniqueness of this mutation, I feel safe in assuming that our most recent common ancestor, my great grandfather, had this same mutation and handed it down. My hope is that someday we will test a participant that has this same mutation. This would most likely indicate a close relationship and could help me find how I connect to the William Barbee line. 

Determine approximately when the mutation occurred. This is done by testing relatives in the same Barbee branch as I described above for the John "Smith" Barbee line. For the second DNA test, I recommend testing the most distant cousin you can find. If the two participants match then your most recent common ancestor likely had the same DNA and the mutation probably occurred in earlier generations. If your DNA results differ, then testing another distant cousin or somewhat closer relative may help narrow when the mutation occurred. 

Rapidly mutating markers may be less useful in identifying how a participant connects to others. Markers that are thought to mutate frequently across the general population are highlighted in red but the actual mutation rates vary by family. For example, almost half of the participants tested that are part of the William Barbee line have a mutation on DYS marker 439. Notice that, of the 3 descendants of Hilary W. Barbee who was born in 1819, 1 participant has a mutation. This mutation likely occurred sometime in the last 4 generations. Even though there are other participants who share the same mutation they have already traced their Barbee lines beyond 1819 when Hilary was born, so this marker is not really helpful in identifying others who may be closely related to the descendant of Hilary Barbee. This mutation may have occurred independently in various branches of the William Barbee line. 

On the other hand, one participant who is descended from Joseph William Barbee who was born in 1864, has a 2 step mutation in DYS marker 439. If another participant is tested who matches this mutation, it could be a significant clue. 

CDYa and CDYb are a special case. They are 2 copies of the same marker that can mutate independently. Unfortunately, there is no way of distinguishing CDYa from CDYb. The lab reports the lowest value for a participant as CDYa and the other value for CDYb. That means that you have to consider those 2 columns together. For example, kit 23700 reports CDYa as 39 and CDYb as 40. However, these results could be reversed, i.e. CDYa might actually be 40 and CDYb might actually be 39. 

Remember that the modal haplotype merely reflects the most common result for those who have been tested. We have no assurance that the modal haplotype reflects the most common value among all Barbees -- it may change as more people are tested. For the markers where everyone in a particular Barbee line matches (like DYS 19 in the William Barbee line), it is likely that William Barbee's son, John (the only son known to have descendants), had the same value for this marker. The more people who are tested the more persuasive this becomes. At this point, we have not tested enough participants in the Andrew Barbee line to reach a similar conclusion. 

Keep in mind that mutations can occur up or down and reverse themselves in subsequent generations. 

All but one of the five documented descendants of William Barbee have mutations. However, without testing other documented descendants of that branch, we have no way of knowing how recently that mutation occurred. It could have been anytime in about the last 9 generations back to William Barbee's son, John. Therefore, be careful about drawing conclusions that you descend from the same line because you share a mutation with one of the documented descendants.
Barbee DNA table is available in the Internet: http://crossroads.barbees.org/DNA/Results.html
Ysearch.org/TTKAK reports a John Barbee b about 1815 in Orange Co. NC and George Washington Barbee b 4 Mar 1845