Blakey and Jackson say Kittles must take pains not to offer genetic information out of context. Jackson is particularly displeased by the fact that Kittles, who reportedly has backing from private investors, plans to charge fees.

Newspaper reports have said he would charge $300, but Kittles told a Chicago radio station Monday the costs would be higher. Jackson scoffs: "Three hundred dollars buys you the sequence of one stretch of DNA." That's about as meaningful as reading only three pages of "War and Peace." "That's fine if somebody thinks they're Yoruba and you match them with Yoruba," Jackson says. "What happens if they don't match up with any of your samples?"

The sole purpose for charging a fee, Kittles said through a Howard University spokeswoman, "would be to cover costs, to conduct further research and protect the integrity of the work and the proper use of the data -- apparently the exact issues about which Dr. Blakey and others seem to be concerned."

Kittles is well aware that his results would offer limited information about one's past. But, the spokeswoman added, "even a hint of their African roots is more than most African-Americans have ever hoped for."

The technology employed by Kittles, amped up by supercomputers in the past few years, can compare stretches of DNA from sites on the Y chromosome and on mitochondrial DNA.

The Y chromosome is passed down by fathers; Y-chromosome tests were used to support the belief that Thomas Jefferson fathered Sally Hemings' child. Some of Kittles' published work involves Y-chromosome inquiries into the history of Finland.

Mitochondrial DNA is inherited from the mother. Mutations show a person's "age" relative to "Eve," theorized to be our common African mother who lived 200,000 years ago. Arizona University professor Michael Hammer used tests of this sort to confirm the Jewish "ethnicity" of the Lemba, a southern African people linked by oral history to one of the lost tribes of Israel.

For some people, such as New York writer Pearl Duncan, DNA seems to offer confirmation of links suggested by language, culture and history.

Duncan was born in the Blue Mountains of Jamaica, where her parents lived on a farm established by rebellious slaves in the 18th century.

A Smithsonian Institution anthropologist told Duncan that some of the odd phrases she'd learned from her parents were Twee language. That brought her to a region of Ghana. DNA tests, conducted with Hammer's assistance, confirmed these were her people.

But Duncan, who is seeking a publisher for a book she has written about her quest, seems to have a straightforward pedigree. That's not the story for many American blacks.

Even if Kittles' tests enabled these African-Americans to identify where one male and one female ancestor came from, what would that reveal about the family tree?

"What people forget is that everyone has two parents and four grandparents and the number of family lines doubles with each generation," says Tony Burroughs, a Chicago genealogist. "I've traced my genealogy back seven generations to my fourth great-grandfather, born in 1781. There are 64 family lines back then and we're still in America. I could have relatives all over Africa as well as South America, Europe and Native Americans."

Kittles' project is, of course, part of a larger social phenomenon, a genealogical craze driven by equal parts Internet and genome project, with a smattering of social disintegration thrown in. Scientists in England have begun offering similar services, including one that offers to link people, via mitochondrial DNA, to one of 12 offspring of the original Eve.

"There are lots of people interested in constructing for themselves some sort of genealogical lineage. What the hell it's going to do for them I don't know," says Clay Dillingham of the Institute of Genetics Education in Santa Fe, N.M. "They're looking for a sense of belonging and family and these companies offer it to you -- even if it's with your australopithecine grandmother."

Some cultural anthropologists are even more dismissive of the genetic approach.

"Culture is not genetically coded," says Richard Wilk, an Indiana University professor who has worked for years with Maya Indians in Belize. In Rwanda, he notes, it was only during the colonial era that the Hutus and Tutsis developed the ironclad sense of difference that contributed to the slaughter of 2 million people a few years ago.

"The example of Rwanda shows you that in 50 to 60 years' time a program of racism and genetic determinism can actually lead to people believing that they are separate and genetically distinct," he says.

And then, he says, there's the "Shirley MacLaine syndrome. Everyone wants to find the king in their ancestry, or the big chief. There's a market for this -- especially among people who want to fill a void that can't be filled for lack of historical records."

But Sam Ford, a Washington television reporter who happened upon Kittles last September while doing a first-person "Roots" investigation, felt that any specific African identity was better than none. When Kittles tested Ford's Y-chromosome, he turned up "hits" in Ibaden, Nigeria.

"OK," Ford acknowledges, "so maybe the Nigerian samples were from people who really came from Ghana. And it's only one branch of my family. But I'm looking for something that gives me some idea of where I come, any idea. This research has a scientific basis. I can't think of anything better.

"Would I bet my life on it? No."