Leaving BYU also gave Sorenson free rein to develop the business side of his genealogy project. In 2001 he had started Relative Genetics as a for-profit company to handle the samples coming in from the Sorenson foundation. Within months, it started selling testing services to the public, looking for business among those curious about their ancestors.

The more data that the Sorenson researchers gather, the more they crave. "We started out believing that 100,000 samples would give us a good cross-section of the world," says Woodward. Now the foundation's latest forecasts call for 500,000 within five years. It is also beginning to expand into mitochondrial DNA, the kind that traces the female line.

Customers who get their Y chromosome tested at Relative Genetics can log on to the Sorenson foundation's Web site and find out for no additional charge what other families have similar genetic markers. Disclosures are most extensive for likely matches with people born at least 100 years ago. For example, the database might show that a man is closely related to a man of the same surname born in England in 1860. Because of a rule barring release of detailed information for people born in the past 100 years, Web-site visitors can't get names and phone numbers of living people who might be distant cousins. Eventually, the foundation may set up ways for limited contacts to occur if all parties want them.

Other research centers are growing fast, too. Family Tree DNA has gathered 31,000 samples so far, and it groups this data into 8,000 surname files so that amateur genealogists can figure out how they relate to people with the same or similar surnames. "If someone's last name is Mauch, they can look at Mock, Mok, Mauck and plenty of other variants in our database," says Bennett Greenspan, chief executive of the Houston company. Sorenson site users are guided only toward the best genetic matches, without the same freedom to probe many surnames.

The new National Geographic-IBM venture has a more academic focus. It is led by Spencer Wells, a leading advocate of the theory that all modern humans descend from a small group that lived in Africa about 60,000 years ago. Wells hopes his data will fill in large blanks in the history of human migrations since then. People who send DNA samples to National Geographic won't be able to connect with specific relatives; they'll only get a general sense of the path their ancestors took in the last 60,000 years. (Separately from its regular business line, Family Tree will conduct the DNA tests on National Geographic's behalf.)