In a search for their ancestors, more than 140 people with variations of the last name Kincaid have taken DNA tests and shared their results on the Internet.
They have found war heroes, sailors and survivors of the Irish potato famine.
They have also stumbled upon bastards, liars and two-timers.
Much of it is ancient history, long-dead ancestors whose dalliances are part of the intrigue of amateur genealogy. But sometimes the findings strike closer to home.
In one case, two brothers were surprised to discover they had different fathers. They confronted their elderly mother, who denied the most obvious possibilities -- that she had been unfaithful to her husband, the man they had always known as Dad, or that one son was adopted.
"It has been traumatic for some to discover their true lineage through the DNA tests," said Don Kincaid, a 76-year-old Texan who oversees the Kincaid surname project and witnessed the brothers' ordeal.
As genetic testing becomes more widespread for medical information, forensics and ancestral research, more people are accidentally uncovering family secrets. Among the most painful are so-called "non-paternity events," cases in which Dad turns out to be someone else.
"It's going to be more and more of a problem," said Dr. Eric Topol, chief of genomic medicine at Scripps Research Institute in La Jolla. Increasing numbers of people will be asking their spouses and parents: "What happened 25 years ago?"
The direct-to-consumer DNA industry sometimes warns customers of the possibility of unintended consequences. But company involvement stops there.
The two Kincaid brothers declined through a spokesman to talk about their experience, calling it too painful.
Others, with the benefit of genetic distance, are more philosophical.
"I'm sure in the history of the Kinkaide family, there's been some fooling around," said 66-year-old Perry Kinkaide.
"If that's unique to this family, I'd be surprised."
Values and behavior
How many of us are not our fathers' children?
The question has fascinated researchers as a window into the gap between a society's stated values and its behavior. A 2005 analysis of 17 studies -- based on blood and DNA tests of various groups -- concluded that the answer varies depending on country and culture. But the average rate is 4%.
The issue has long lurked in the background of medicine. It's not hard to figure out if your blood type is compatible with Mom's and Dad's(If they are both A positive and you are B positive, you have a problem.)
A recent survey of 56 kidney transplant centers by the University of Maryland showed that 70% had stumbled upon at least one case of non-paternity as a result of testing potential organ donors.
DNA testing has opened the gates of possibility. The potential for surprises exists whenever members of the same family are tested.
For example, researchers looking for the genetic fingerprints of certain diseases have long compared child and parent DNA.
Every so often, mismatches pop up that raise the possibility of hanky-panky.
In research, subjects have signed waivers agreeing that discoveries of non-paternity will not be revealed to them or anybody else.
But in medical practice, the truth has a way of cruelly asserting itself.
In the most common scenario, a child is born with Tay-Sachs disease, cystic fibrosis or another disorder that requires the contribution of a certain gene from each parent. The parents are tested, and the father is found not to carry the gene.
Breaking the news falls to genetic counselors, who often must balance competing ethical imperatives.
"Non-paternity is one of the issues that genetic counselors dread but at some point in their careers will have to deal with," said Andrea Atherton, a counselor at Children's Mercy Hospital in Kansas City, Mo.
Standard practice is to tell only the mother, who usually already suspects it, genetic counselors say.
Ana Morales, a genetic counselor at the University of Miami, recalled the case of a child diagnosed with a type of albinism that can be accompanied by lung and kidney disease.
The mother "told me she was having an affair," Morales said.
"She said she would be in physical danger [if her husband found out]. He had threatened her if she was unfaithful."
Morales did not tell him.
But withholding the information means that the woman's husband lives with the false belief that he is a carrier of a genetic disorder.
That sort of information is far from benign, said Dr. Wayne Grody, a UCLA geneticist. It could convince him to give up on the idea of having children. And in the event that the wife becomes pregnant by her husband, perpetuating the lie could require unnecessary prenatal testing.
"Why would you expose the next fetus to the risk of amniocentesis?" Grody asked.
In the growing world of direct-to-consumer DNA testing, customers are usually on their own to discover and digest non-paternity.
The industry has ballooned to more than three dozen companies from its inception about nine years ago.
There is a wide range, from those that offer basic ancestry testing to a few that scan several hundred thousand genes looking for susceptibility to certain diseases.
Scans, which require a cheek swab or a vial of saliva, cost $100 to $1,000.
23andme, a year-old company that mines genomes for medical information, warns customers in a consent form that testing could reveal that "your father is not genetically your father."
The company allows customers to display vast swaths of their DNA next to that of relatives -- with everyone's consent -- making cases of non-paternity easy to spot, geneticists say.
Rachel Cohen, a spokeswoman, said she is not aware of any such cases, though the company does not look for them or necessarily hear about them from customers.
In ancestry testing, non-paternity shows up most often in comparisons of Y chromosomes, passed from father to son. (Some companies also test mitochondrial DNA, which everybody inherits from their mothers, but it is less useful in genealogy because surnames are usually passed down the male line. Women often submit the Y-DNA of a close male relative.)
All the males in a single bloodline have the same Y-chromosome, except for the tiny mutations that accumulate over generations.
If a father and son have vastly different Y-chromosomes, they are not related.
If cousins have dramatically different Y-chromosomes, it is safe to conclude that somewhere along the line that joins them, someone is keeping a secret.
And if the Y-chromosome looks nothing like those of the other people with the same last name who have posted theirs on the Internet, it is fair to wonder whether somebody is hiding something. Of the 147 people in the Kincaid project, most fall into four main groups of Y-DNA.
But about 10 Kincaids didn't match up with anybody else.
"You can let your mind run wild," said Bob Kinkaid, 68, from Star Tannery, Va., who didn't find any Kincaids with Y-DNA similar to his.
"You never know when a male child may have been adopted," he offered.
Perry Kinkaide, who lives in Edmonton, Canada, said that after two decades of tracing his family's paper trail, he thought he knew many of his ancestors.
Then he sent in a DNA sample. The results suggest that he wasn't biologically related to the people he had been studying -- not that it bothered him.
Any indiscretions probably happened at least a few generations back, he guessed.
"I looked like my father," he said. "We even had the same walk."
Don Severs, a 47-year-old data manager from Des Moines, Iowa, said DNA helped him confirm that his great-great-grandfather, William Severs, born in 1815, was not a Severs at all.
William Severs' biological father was a postmaster named George Kinkade, and his mother the family's housekeeper. When she became pregnant, the family arranged for her to marry a Severs.
The secret probably saved his mother and the Kinkades from scandal.
Don Severs, who has the Y-DNA of the Kincaid clan, wonders if anybody would go through such pains today.
"Now illegitimacy is no big deal," he said.