Woodward soon embraced the idea of building a database that could rival anything in academia. After Sorenson got back from his trip to Norway in 1999, the two settled on a plan for the Sorenson foundation. It would collect small samples of many populations, focusing on the Y chromosome. Once typical profiles for each one were established, people could plug in their own data and figure out what region their forebears along the male line might have come from.

The idea didn't stop there. Woodward decided to get detailed genealogies from those who contributed their DNA. Eventually, he figured, some people would be able to submit their DNA sample and find a connection not just to a region but also to a specific ancestor. To protect donors' privacy, the Sorenson team decided not to release details about anyone born in the past 100 years.

In early 2000, notices began appearing on BYU's main campus in Provo, inviting students to visit Woodward's office so they could contribute blood samples and at least four generations of family history, including birthdates and country of origin. Unsure whether anyone would bother, Woodward offered participants $10 apiece for their trouble.

Right away, Woodward's office was mobbed. Seventy students hovered outside the door at 9 a.m. on the first sampling day, March 6, 2000. It took eight hours to process them all. That stampede repeated itself daily for months.

While BYU students' enthusiasm helped generate thousands of samples in the first few months, it also left the project overweighted with people of Anglo-Saxon, Germanic or Scandinavian origin. Seeking greater diversity, one of Woodward's graduate students, Ugo Perego, pored through Census Bureau data in search of unusual immigrant enclaves in the United States.

Starting in 2001, Perego visited towns such as Red Lodge, Mont., to sample people of Finnish origin; Malad City, Idaho, for Welsh-Americans; and New Bedford, Mass., for Portuguese-Americans. To get past airport security with his attache cases full of test tubes, Perego recalls that he had to get a letter from the Centers for Disease Control.

When a Mexican politician visiting Utah offhandedly suggested studying the indigenous Totonac population, Sorenson researchers rushed days later to a remote corner of eastern Mexico. No matter that researchers hadn't heard of the Totonac until then. They returned with more than 100 DNA samples.

Most of the Sorenson effort overseas targets countries of ancestral significance to many Americans, such as Nigeria and China. Sorenson scientists also decided in 2003 to emulate other geneticists' adoption of a painless alternative to blood sampling. A brisk mouthwash rinse, it turned out, could collect enough cheek cells to generate reliable lab results.

In 2003, the project severed its ties with BYU and relocated to Sorenson's corporate headquarters in Salt Lake City. The university was running out of lab space, and the switch helped allay any concerns among non-Mormons that the project might have a religious agenda.